!pip install --upgrade plotly
!pip install jupyter-dash
!pip install cufflinks --upgrade
!pip install ftfy
!pip install langdetect
# !pip install googletrans
!pip install goslate

Requirement already up-to-date: plotly in /usr/local/lib/python3.7/dist-packages (4.14.3)
Requirement already satisfied, skipping upgrade: retrying>=1.3.3 in /usr/local/lib/python3.7/dist-packages (from plotly) (1.3.3)
Requirement already satisfied, skipping upgrade: six in /usr/local/lib/python3.7/dist-packages (from plotly) (1.15.0)
Requirement already satisfied: jupyter-dash in /usr/local/lib/python3.7/dist-packages (0.4.0)
Requirement already satisfied: ipython in /usr/local/lib/python3.7/dist-packages (from jupyter-dash) (5.5.0)
Requirement already satisfied: requests in /usr/local/lib/python3.7/dist-packages (from jupyter-dash) (2.23.0)
Requirement already satisfied: dash in /usr/local/lib/python3.7/dist-packages (from jupyter-dash) (1.20.0)
Requirement already satisfied: ipykernel in /usr/local/lib/python3.7/dist-packages (from jupyter-dash) (4.10.1)
Requirement already satisfied: flask in /usr/local/lib/python3.7/dist-packages (from jupyter-dash) (1.1.4)
Requirement already satisfied: ansi2html in /usr/local/lib/python3.7/dist-packages (from jupyter-dash) (1.6.0)
Requirement already satisfied: retrying in /usr/local/lib/python3.7/dist-packages (from jupyter-dash) (1.3.3)
Requirement already satisfied: traitlets>=4.2 in /usr/local/lib/python3.7/dist-packages (from ipython->jupyter-dash) (5.0.5)
Requirement already satisfied: simplegeneric>0.8 in /usr/local/lib/python3.7/dist-packages (from ipython->jupyter-dash) (0.8.1)
Requirement already satisfied: prompt-toolkit<2.0.0,>=1.0.4 in /usr/local/lib/python3.7/dist-packages (from ipython->jupyter-dash) (1.0.18)
Requirement already satisfied: pickleshare in /usr/local/lib/python3.7/dist-packages (from ipython->jupyter-dash) (0.7.5)
Requirement already satisfied: decorator in /usr/local/lib/python3.7/dist-packages (from ipython->jupyter-dash) (4.4.2)
Requirement already satisfied: pexpect; sys_platform != "win32" in /usr/local/lib/python3.7/dist-packages (from ipython->jupyter-dash) (4.8.0)
Requirement already satisfied: setuptools>=18.5 in /usr/local/lib/python3.7/dist-packages (from ipython->jupyter-dash) (57.0.0)
Requirement already satisfied: pygments in /usr/local/lib/python3.7/dist-packages (from ipython->jupyter-dash) (2.6.1)
Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests->jupyter-dash) (3.0.4)
Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests->jupyter-dash) (1.24.3)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests->jupyter-dash) (2020.12.5)
Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests->jupyter-dash) (2.10)
Requirement already satisfied: future in /usr/local/lib/python3.7/dist-packages (from dash->jupyter-dash) (0.16.0)
Requirement already satisfied: dash-html-components==1.1.3 in /usr/local/lib/python3.7/dist-packages (from dash->jupyter-dash) (1.1.3)
Requirement already satisfied: flask-compress in /usr/local/lib/python3.7/dist-packages (from dash->jupyter-dash) (1.9.0)
Requirement already satisfied: dash-table==4.11.3 in /usr/local/lib/python3.7/dist-packages (from dash->jupyter-dash) (4.11.3)
Requirement already satisfied: dash-core-components==1.16.0 in /usr/local/lib/python3.7/dist-packages (from dash->jupyter-dash) (1.16.0)
Requirement already satisfied: dash-renderer==1.9.1 in /usr/local/lib/python3.7/dist-packages (from dash->jupyter-dash) (1.9.1)
Requirement already satisfied: plotly in /usr/local/lib/python3.7/dist-packages (from dash->jupyter-dash) (4.14.3)
Requirement already satisfied: tornado>=4.0 in /usr/local/lib/python3.7/dist-packages (from ipykernel->jupyter-dash) (5.1.1)
Requirement already satisfied: jupyter-client in /usr/local/lib/python3.7/dist-packages (from ipykernel->jupyter-dash) (5.3.5)
Requirement already satisfied: Jinja2<3.0,>=2.10.1 in /usr/local/lib/python3.7/dist-packages (from flask->jupyter-dash) (2.11.3)
Requirement already satisfied: Werkzeug<2.0,>=0.15 in /usr/local/lib/python3.7/dist-packages (from flask->jupyter-dash) (1.0.1)
Requirement already satisfied: click<8.0,>=5.1 in /usr/local/lib/python3.7/dist-packages (from flask->jupyter-dash) (7.1.2)
Requirement already satisfied: itsdangerous<2.0,>=0.24 in /usr/local/lib/python3.7/dist-packages (from flask->jupyter-dash) (1.1.0)
Requirement already satisfied: six>=1.7.0 in /usr/local/lib/python3.7/dist-packages (from retrying->jupyter-dash) (1.15.0)
Requirement already satisfied: ipython-genutils in /usr/local/lib/python3.7/dist-packages (from traitlets>=4.2->ipython->jupyter-dash) (0.2.0)
Requirement already satisfied: wcwidth in /usr/local/lib/python3.7/dist-packages (from prompt-toolkit<2.0.0,>=1.0.4->ipython->jupyter-dash) (0.2.5)
Requirement already satisfied: ptyprocess>=0.5 in /usr/local/lib/python3.7/dist-packages (from pexpect; sys_platform != "win32"->ipython->jupyter-dash) (0.7.0)
Requirement already satisfied: brotli in /usr/local/lib/python3.7/dist-packages (from flask-compress->dash->jupyter-dash) (1.0.9)
Requirement already satisfied: jupyter-core>=4.6.0 in /usr/local/lib/python3.7/dist-packages (from jupyter-client->ipykernel->jupyter-dash) (4.7.1)
Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.7/dist-packages (from jupyter-client->ipykernel->jupyter-dash) (2.8.1)
Requirement already satisfied: pyzmq>=13 in /usr/local/lib/python3.7/dist-packages (from jupyter-client->ipykernel->jupyter-dash) (22.0.3)
Requirement already satisfied: MarkupSafe>=0.23 in /usr/local/lib/python3.7/dist-packages (from Jinja2<3.0,>=2.10.1->flask->jupyter-dash) (2.0.1)
Requirement already up-to-date: cufflinks in /usr/local/lib/python3.7/dist-packages (0.17.3)
Requirement already satisfied, skipping upgrade: numpy>=1.9.2 in /usr/local/lib/python3.7/dist-packages (from cufflinks) (1.19.5)
Requirement already satisfied, skipping upgrade: colorlover>=0.2.1 in /usr/local/lib/python3.7/dist-packages (from cufflinks) (0.3.0)
Requirement already satisfied, skipping upgrade: plotly>=4.1.1 in /usr/local/lib/python3.7/dist-packages (from cufflinks) (4.14.3)
Requirement already satisfied, skipping upgrade: six>=1.9.0 in /usr/local/lib/python3.7/dist-packages (from cufflinks) (1.15.0)
Requirement already satisfied, skipping upgrade: ipywidgets>=7.0.0 in /usr/local/lib/python3.7/dist-packages (from cufflinks) (7.6.3)
Requirement already satisfied, skipping upgrade: setuptools>=34.4.1 in /usr/local/lib/python3.7/dist-packages (from cufflinks) (57.0.0)
Requirement already satisfied, skipping upgrade: pandas>=0.19.2 in /usr/local/lib/python3.7/dist-packages (from cufflinks) (1.1.5)
Requirement already satisfied, skipping upgrade: ipython>=5.3.0 in /usr/local/lib/python3.7/dist-packages (from cufflinks) (5.5.0)
Requirement already satisfied, skipping upgrade: retrying>=1.3.3 in /usr/local/lib/python3.7/dist-packages (from plotly>=4.1.1->cufflinks) (1.3.3)
Requirement already satisfied, skipping upgrade: widgetsnbextension~=3.5.0 in /usr/local/lib/python3.7/dist-packages (from ipywidgets>=7.0.0->cufflinks) (3.5.1)
Requirement already satisfied, skipping upgrade: ipykernel>=4.5.1 in /usr/local/lib/python3.7/dist-packages (from ipywidgets>=7.0.0->cufflinks) (4.10.1)
Requirement already satisfied, skipping upgrade: nbformat>=4.2.0 in /usr/local/lib/python3.7/dist-packages (from ipywidgets>=7.0.0->cufflinks) (5.1.3)
Requirement already satisfied, skipping upgrade: traitlets>=4.3.1 in /usr/local/lib/python3.7/dist-packages (from ipywidgets>=7.0.0->cufflinks) (5.0.5)
Requirement already satisfied, skipping upgrade: jupyterlab-widgets>=1.0.0; python_version >= "3.6" in /usr/local/lib/python3.7/dist-packages (from ipywidgets>=7.0.0->cufflinks) (1.0.0)
Requirement already satisfied, skipping upgrade: python-dateutil>=2.7.3 in /usr/local/lib/python3.7/dist-packages (from pandas>=0.19.2->cufflinks) (2.8.1)
Requirement already satisfied, skipping upgrade: pytz>=2017.2 in /usr/local/lib/python3.7/dist-packages (from pandas>=0.19.2->cufflinks) (2018.9)
Requirement already satisfied, skipping upgrade: pexpect; sys_platform != "win32" in /usr/local/lib/python3.7/dist-packages (from ipython>=5.3.0->cufflinks) (4.8.0)
Requirement already satisfied, skipping upgrade: decorator in /usr/local/lib/python3.7/dist-packages (from ipython>=5.3.0->cufflinks) (4.4.2)
Requirement already satisfied, skipping upgrade: simplegeneric>0.8 in /usr/local/lib/python3.7/dist-packages (from ipython>=5.3.0->cufflinks) (0.8.1)
Requirement already satisfied, skipping upgrade: pygments in /usr/local/lib/python3.7/dist-packages (from ipython>=5.3.0->cufflinks) (2.6.1)
Requirement already satisfied, skipping upgrade: prompt-toolkit<2.0.0,>=1.0.4 in /usr/local/lib/python3.7/dist-packages (from ipython>=5.3.0->cufflinks) (1.0.18)
Requirement already satisfied, skipping upgrade: pickleshare in /usr/local/lib/python3.7/dist-packages (from ipython>=5.3.0->cufflinks) (0.7.5)
Requirement already satisfied, skipping upgrade: notebook>=4.4.1 in /usr/local/lib/python3.7/dist-packages (from widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (5.3.1)
Requirement already satisfied, skipping upgrade: tornado>=4.0 in /usr/local/lib/python3.7/dist-packages (from ipykernel>=4.5.1->ipywidgets>=7.0.0->cufflinks) (5.1.1)
Requirement already satisfied, skipping upgrade: jupyter-client in /usr/local/lib/python3.7/dist-packages (from ipykernel>=4.5.1->ipywidgets>=7.0.0->cufflinks) (5.3.5)
Requirement already satisfied, skipping upgrade: ipython-genutils in /usr/local/lib/python3.7/dist-packages (from nbformat>=4.2.0->ipywidgets>=7.0.0->cufflinks) (0.2.0)
Requirement already satisfied, skipping upgrade: jupyter-core in /usr/local/lib/python3.7/dist-packages (from nbformat>=4.2.0->ipywidgets>=7.0.0->cufflinks) (4.7.1)
Requirement already satisfied, skipping upgrade: jsonschema!=2.5.0,>=2.4 in /usr/local/lib/python3.7/dist-packages (from nbformat>=4.2.0->ipywidgets>=7.0.0->cufflinks) (2.6.0)
Requirement already satisfied, skipping upgrade: ptyprocess>=0.5 in /usr/local/lib/python3.7/dist-packages (from pexpect; sys_platform != "win32"->ipython>=5.3.0->cufflinks) (0.7.0)
Requirement already satisfied, skipping upgrade: wcwidth in /usr/local/lib/python3.7/dist-packages (from prompt-toolkit<2.0.0,>=1.0.4->ipython>=5.3.0->cufflinks) (0.2.5)
Requirement already satisfied, skipping upgrade: nbconvert in /usr/local/lib/python3.7/dist-packages (from notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (5.6.1)
Requirement already satisfied, skipping upgrade: terminado>=0.8.1 in /usr/local/lib/python3.7/dist-packages (from notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (0.10.0)
Requirement already satisfied, skipping upgrade: Send2Trash in /usr/local/lib/python3.7/dist-packages (from notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (1.5.0)
Requirement already satisfied, skipping upgrade: jinja2 in /usr/local/lib/python3.7/dist-packages (from notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (2.11.3)
Requirement already satisfied, skipping upgrade: pyzmq>=13 in /usr/local/lib/python3.7/dist-packages (from jupyter-client->ipykernel>=4.5.1->ipywidgets>=7.0.0->cufflinks) (22.0.3)
Requirement already satisfied, skipping upgrade: mistune<2,>=0.8.1 in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (0.8.4)
Requirement already satisfied, skipping upgrade: bleach in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (3.3.0)
Requirement already satisfied, skipping upgrade: defusedxml in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (0.7.1)
Requirement already satisfied, skipping upgrade: testpath in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (0.5.0)
Requirement already satisfied, skipping upgrade: pandocfilters>=1.4.1 in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (1.4.3)
Requirement already satisfied, skipping upgrade: entrypoints>=0.2.2 in /usr/local/lib/python3.7/dist-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (0.3)
Requirement already satisfied, skipping upgrade: MarkupSafe>=0.23 in /usr/local/lib/python3.7/dist-packages (from jinja2->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (2.0.1)
Requirement already satisfied, skipping upgrade: packaging in /usr/local/lib/python3.7/dist-packages (from bleach->nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (20.9)
Requirement already satisfied, skipping upgrade: webencodings in /usr/local/lib/python3.7/dist-packages (from bleach->nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (0.5.1)
Requirement already satisfied, skipping upgrade: pyparsing>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from packaging->bleach->nbconvert->notebook>=4.4.1->widgetsnbextension~=3.5.0->ipywidgets>=7.0.0->cufflinks) (2.4.7)
Requirement already satisfied: ftfy in /usr/local/lib/python3.7/dist-packages (6.0.3)
Requirement already satisfied: wcwidth in /usr/local/lib/python3.7/dist-packages (from ftfy) (0.2.5)
Requirement already satisfied: langdetect in /usr/local/lib/python3.7/dist-packages (1.0.9)
Requirement already satisfied: six in /usr/local/lib/python3.7/dist-packages (from langdetect) (1.15.0)
Requirement already satisfied: goslate in /usr/local/lib/python3.7/dist-packages (1.5.1)
Requirement already satisfied: futures in /usr/local/lib/python3.7/dist-packages (from goslate) (3.1.1)

import warnings
warnings.filterwarnings('ignore') 

# Utilities
from time import time
import os, sys, itertools, re
import warnings, string
from ftfy import fix_encoding, fix_text, badness
from wordcloud import WordCloud, STOPWORDS, ImageColorGenerator

# Numerical calculation
import numpy as np

# Data Handling
import pandas as pd

# Data Visualization
import matplotlib.pyplot as plt
from matplotlib.ticker import PercentFormatter
import seaborn as sns
import cufflinks as cf
import plotly as py
import plotly.express as px
import plotly.graph_objs as go
from plotly import __version__
from plotly.offline import download_plotlyjs, init_notebook_mode, plot, iplot
from IPython.display import HTML

init_notebook_mode(connected=False)
cf.go_offline()
%matplotlib inline

#NLP
import string
import re
from collections import Counter
from langdetect import detect
import nltk
from nltk.corpus import stopwords
from nltk import tokenize
from textblob import Word
from wordcloud import WordCloud, STOPWORDS 

# Tools & Evaluation metrics
from scipy.stats import zscore
from sklearn import preprocessing
from sklearn.utils.class_weight import compute_class_weight
from sklearn.preprocessing import label_binarize
from sklearn.decomposition import LatentDirichletAllocation
from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer, TfidfTransformer
from sklearn.metrics import roc_curve, roc_auc_score, auc, accuracy_score, confusion_matrix, f1_score, classification_report, precision_recall_curve, precision_recall_fscore_support
from sklearn.model_selection import train_test_split

# Traditional Modeling
from sklearn.multiclass import OneVsRestClassifier
from sklearn.linear_model import LogisticRegression
from sklearn.linear_model import SGDClassifier
from sklearn.naive_bayes import MultinomialNB, BernoulliNB
from sklearn.neighbors import KNeighborsClassifier
from sklearn.pipeline import Pipeline
from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer, TfidfTransformer
from sklearn.svm import SVC, LinearSVC
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier
import xgboost as xgb

# Deep learning models
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Activation, Dropout
from tensorflow.keras.optimizers import SGD

# Translation APIs
from goslate import Goslate # Provided by Google

def configure_plotly_browser_state():
  import IPython
  display(IPython.core.display.HTML('''
        <script src="/static/components/requirejs/require.js"></script>
        <script>
          requirejs.config({
            paths: {
              base: '/static/base',
              plotly: 'https://cdn.plot.ly/plotly-1.5.1.min.js?noext',
            },
          });
        </script>
        '''))

nltk.download('stopwords')
nltk.download('wordnet')

[nltk_data] Downloading package stopwords to /root/nltk_data...
[nltk_data]   Package stopwords is already up-to-date!
[nltk_data] Downloading package wordnet to /root/nltk_data...
[nltk_data]   Package wordnet is already up-to-date!

True

 from google.colab import drive
drive.mount('/content/drive/')

Drive already mounted at /content/drive/; to attempt to forcibly remount, call drive.mount("/content/drive/", force_remount=True).

project_path =  ('/content/drive/My Drive/Colab Notebooks/Capstone_Project_AIML/')
file_name ='input_data.xlsx'

ticket_df = pd.read_excel(project_path+file_name)
ticket_df.head(5)

	Short description	Description	Caller	Assignment group
0	login issue	-verified user details.(employee# & manager na...	spxjnwir pjlcoqds	GRP_0
1	outlook	\r\n\r\nreceived from: hmjdrvpb.komuaywn@gmail...	hmjdrvpb komuaywn	GRP_0
2	cant log in to vpn	\r\n\r\nreceived from: eylqgodm.ybqkwiam@gmail...	eylqgodm ybqkwiam	GRP_0
3	unable to access hr_tool page	unable to access hr_tool page	xbkucsvz gcpydteq	GRP_0
4	skype error	skype error	owlgqjme qhcozdfx	GRP_0

ticket_df.describe()

	Short description	Description	Caller	Assignment group
count	8492	8499	8500	8500
unique	7481	7817	2950	74
top	password reset	the	bpctwhsn kzqsbmtp	GRP_0
freq	38	56	810	3976

ticket_df['Assignment group'].value_counts()

GRP_0     3976
GRP_8      661
GRP_24     289
GRP_12     257
GRP_9      252
          ... 
GRP_73       1
GRP_64       1
GRP_70       1
GRP_67       1
GRP_61       1
Name: Assignment group, Length: 74, dtype: int64

ticket_df['Assignment group'].unique()

array(['GRP_0', 'GRP_1', 'GRP_3', 'GRP_4', 'GRP_5', 'GRP_6', 'GRP_7',
       'GRP_8', 'GRP_9', 'GRP_10', 'GRP_11', 'GRP_12', 'GRP_13', 'GRP_14',
       'GRP_15', 'GRP_16', 'GRP_17', 'GRP_18', 'GRP_19', 'GRP_2',
       'GRP_20', 'GRP_21', 'GRP_22', 'GRP_23', 'GRP_24', 'GRP_25',
       'GRP_26', 'GRP_27', 'GRP_28', 'GRP_29', 'GRP_30', 'GRP_31',
       'GRP_33', 'GRP_34', 'GRP_35', 'GRP_36', 'GRP_37', 'GRP_38',
       'GRP_39', 'GRP_40', 'GRP_41', 'GRP_42', 'GRP_43', 'GRP_44',
       'GRP_45', 'GRP_46', 'GRP_47', 'GRP_48', 'GRP_49', 'GRP_50',
       'GRP_51', 'GRP_52', 'GRP_53', 'GRP_54', 'GRP_55', 'GRP_56',
       'GRP_57', 'GRP_58', 'GRP_59', 'GRP_60', 'GRP_61', 'GRP_32',
       'GRP_62', 'GRP_63', 'GRP_64', 'GRP_65', 'GRP_66', 'GRP_67',
       'GRP_68', 'GRP_69', 'GRP_70', 'GRP_71', 'GRP_72', 'GRP_73'],
      dtype=object)

len(ticket_df['Assignment group'].unique())

74

Dataset contains 74 unique Assignment Groups

tkt_df = ticket_df['Assignment group'].value_counts().reset_index()
tkt_df['percentage'] = (tkt_df['Assignment group']/tkt_df['Assignment group'].sum())*100
tkt_df.head()

	index	Assignment group	percentage
0	GRP_0	3976	46.776471
1	GRP_8	661	7.776471
2	GRP_24	289	3.400000
3	GRP_12	257	3.023529
4	GRP_9	252	2.964706

EXPLORATORY DATA ANALYSIS

Basic Data Cleaning

ticket_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8500 entries, 0 to 8499
Data columns (total 4 columns):
 #   Column             Non-Null Count  Dtype 
---  ------             --------------  ----- 
 0   Short description  8492 non-null   object
 1   Description        8499 non-null   object
 2   Caller             8500 non-null   object
 3   Assignment group   8500 non-null   object
dtypes: object(4)
memory usage: 265.8+ KB

ticket_df.isnull().sum()

Short description    8
Description          1
Caller               0
Assignment group     0
dtype: int64

There are null values in columns - 'Short description' & 'Description'

ticket_df[pd.isnull(ticket_df).any(axis=1)]

	Short description	Description	Caller	Assignment group
2604	NaN	\r\n\r\nreceived from: ohdrnswl.rezuibdt@gmail...	ohdrnswl rezuibdt	GRP_34
3383	NaN	\r\n-connected to the user system using teamvi...	qftpazns fxpnytmk	GRP_0
3906	NaN	-user unable tologin to vpn.\r\n-connected to...	awpcmsey ctdiuqwe	GRP_0
3910	NaN	-user unable tologin to vpn.\r\n-connected to...	rhwsmefo tvphyura	GRP_0
3915	NaN	-user unable tologin to vpn.\r\n-connected to...	hxripljo efzounig	GRP_0
3921	NaN	-user unable tologin to vpn.\r\n-connected to...	cziadygo veiosxby	GRP_0
3924	NaN	name:wvqgbdhm fwchqjor\nlanguage:\nbrowser:mic...	wvqgbdhm fwchqjor	GRP_0
4341	NaN	\r\n\r\nreceived from: eqmuniov.ehxkcbgj@gmail...	eqmuniov ehxkcbgj	GRP_0
4395	i am locked out of skype	NaN	viyglzfo ajtfzpkb	GRP_0

Remove null values by converting it to string

# Convert datatype to string
ticket_df = ticket_df.applymap(str)
ticket_df.count()

Short description    8500
Description          8500
Caller               8500
Assignment group     8500
dtype: int64

ticket_df[pd.isnull(ticket_df).any(axis=1)]

	Short description	Description	Caller	Assignment group

ticket_df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 8500 entries, 0 to 8499
Data columns (total 4 columns):
 #   Column             Non-Null Count  Dtype 
---  ------             --------------  ----- 
 0   Short description  8500 non-null   object
 1   Description        8500 non-null   object
 2   Caller             8500 non-null   object
 3   Assignment group   8500 non-null   object
dtypes: object(4)
memory usage: 265.8+ KB

#Concatenate Short Description and Description columns
ticket_df['Full_Description'] = ticket_df['Short description'] + ' ' +ticket_df['Description']
ticket_df.head(15)

	Short description	Description	Caller	Assignment group	Full_Description
0	login issue	-verified user details.(employee# & manager na...	spxjnwir pjlcoqds	GRP_0	login issue -verified user details.(employee# ...
1	outlook	\r\n\r\nreceived from: hmjdrvpb.komuaywn@gmail...	hmjdrvpb komuaywn	GRP_0	outlook \r\n\r\nreceived from: hmjdrvpb.komuay...
2	cant log in to vpn	\r\n\r\nreceived from: eylqgodm.ybqkwiam@gmail...	eylqgodm ybqkwiam	GRP_0	cant log in to vpn \r\n\r\nreceived from: eylq...
3	unable to access hr_tool page	unable to access hr_tool page	xbkucsvz gcpydteq	GRP_0	unable to access hr_tool page unable to access...
4	skype error	skype error	owlgqjme qhcozdfx	GRP_0	skype error skype error
5	unable to log in to engineering tool and skype	unable to log in to engineering tool and skype	eflahbxn ltdgrvkz	GRP_0	unable to log in to engineering tool and skype...
6	event: critical:HostName_221.company.com the v...	event: critical:HostName_221.company.com the v...	jyoqwxhz clhxsoqy	GRP_1	event: critical:HostName_221.company.com the v...
7	ticket_no1550391- employment status - new non-...	ticket_no1550391- employment status - new non-...	eqzibjhw ymebpoih	GRP_0	ticket_no1550391- employment status - new non-...
8	unable to disable add ins on outlook	unable to disable add ins on outlook	mdbegvct dbvichlg	GRP_0	unable to disable add ins on outlook unable to...
9	ticket update on inplant_874773	ticket update on inplant_874773	fumkcsji sarmtlhy	GRP_0	ticket update on inplant_874773 ticket update ...
10	engineering tool says not connected and unable...	engineering tool says not connected and unable...	badgknqs xwelumfz	GRP_0	engineering tool says not connected and unable...
11	hr_tool site not loading page correctly	hr_tool site not loading page correctly	dcqsolkx kmsijcuz	GRP_0	hr_tool site not loading page correctly hr_too...
12	unable to login to hr_tool to sgxqsuojr xwbeso...	unable to login to hr_tool to sgxqsuojr xwbeso...	oblekmrw qltgvspb	GRP_0	unable to login to hr_tool to sgxqsuojr xwbeso...
13	user wants to reset the password	user wants to reset the password	iftldbmu fujslwby	GRP_0	user wants to reset the password user wants to...
14	unable to open payslips	unable to open payslips	epwyvjsz najukwho	GRP_0	unable to open payslips unable to open payslips

#Detect and corrected data with Mojibakes

def is_mojibake_impacted(text):
    if not badness.badness(text):
        # nothing weird, should be okay
        return True
    try:
        text.encode('sloppy-windows-1252')
    except UnicodeEncodeError:
        # Not CP-1252 encodable, probably fine
        return True
    else:
        # Encodable as CP-1252, Mojibake alert level high
        return False
    
# Check the dataset for mojibake impact
ticket_df[~ticket_df.iloc[:,:-1].applymap(is_mojibake_impacted).all(1)]

	Short description	Description	Caller	Assignment group	Full_Description
99	password expiry tomorrow	\n\nreceived from: ecprjbod.litmjwsy@gmail.com...	ecprjbod litmjwsy	GRP_0	password expiry tomorrow \n\nreceived from: ec...
116	server issues	\r\n\r\nreceived from: bgqpotek.cuxakvml@gmail...	bgqpotek cuxakvml	GRP_0	server issues \r\n\r\nreceived from: bgqpotek....
124	mobile device activation	from: tvcdfqgp nrbcqwgj \nsent: friday, octobe...	tvcdfqgp nrbcqwgj	GRP_0	mobile device activation from: tvcdfqgp nrbcqw...
162	access to bex	\r\n\r\nreceived from: yfqoaepn.xnezhsit@gmail...	yfqoaepn xnezhsit	GRP_0	access to bex \r\n\r\nreceived from: yfqoaepn....
164	撤回: ticket_no1564867 -- comments added	\n\nreceived from: abcdri@company.com\n\nwindy...	tycludks cjofwigv	GRP_0	撤回: ticket_no1564867 -- comments added \n\...
...	...	...	...	...	...
8470	please review your recent ticketing_tool ticke...	from: mikhghytr wafglhdrhjop \nsent: thursday,...	azxhejvq fyemlavd	GRP_16	please review your recent ticketing_tool ticke...
8471	电脑开机开不出来	to 小贺，早上电脑开机开不出来	xqyjztnm onfusvlz	GRP_30	电脑开机开不出来 to 小贺，早上电...
8480	customer group enhanced field	\r\n\r\nreceived from: nlearzwi.ukdzstwi@gmail...	nlearzwi ukdzstwi	GRP_9	customer group enhanced field \r\n\r\nreceived...
8498	machine não está funcionando	i am unable to access the machine utilities to...	ufawcgob aowhxjky	GRP_62	machine não está funcionando i am unable to ...
8499	an mehreren pc`s lassen sich verschiedene prgr...	an mehreren pc`s lassen sich verschiedene prgr...	kqvbrspl jyzoklfx	GRP_49	an mehreren pc`s lassen sich verschiedene prgr...

853 rows × 5 columns

ticket_df.iloc[8471,:]

Short description                             电脑开机开不出来
Description                 to 小贺，早上电脑开机开不出来
Caller                                               xqyjztnm onfusvlz
Assignment group                                                GRP_30
Full_Description     电脑开机开不出来 to 小贺，早上电...
Name: 8471, dtype: object

ticket_df['Full_Description'] = ticket_df['Full_Description'].apply(fix_text)

ticket_df.iloc[8471,:]

Short description                      电脑开机开不出来
Description          to 小贺，早上电脑开机开不出来
Caller                                        xqyjztnm onfusvlz
Assignment group                                         GRP_30
Full_Description                      电脑开机开不出来 to 小贺,早上电脑开机开不出来
Name: 8471, dtype: object

The converted mojibakes data seems to be Mandarin. Thus there is presence of other non-english language that need to be converted.

Language detection

def fn_lan_detect(df):                                        
   try:   
      # print("lang: ", detect(df))                                                       
      return detect(df)                                      
   except:                                                       
      return 'no'                                                  

ticket_df['language'] = ticket_df['Full_Description'].apply(fn_lan_detect)

ticket_df.iloc[8471,:]

Short description                      电脑开机开不出来
Description          to 小贺，早上电脑开机开不出来
Caller                                        xqyjztnm onfusvlz
Assignment group                                         GRP_30
Full_Description                      电脑开机开不出来 to 小贺,早上电脑开机开不出来
language                                                  zh-cn
Name: 8471, dtype: object

ticket_df["language"].value_counts()

en       7094
de        414
af        264
it        121
fr        112
nl         61
no         61
sv         55
es         50
zh-cn      50
ca         45
pl         30
da         30
pt         19
ko         16
tl         11
et         11
ro         10
cy          9
hr          7
sq          6
id          5
fi          4
sl          4
so          3
vi          2
lt          2
cs          2
lv          1
tr          1
Name: language, dtype: int64

lang_count = ticket_df["language"].value_counts().reset_index()
lang_count

	index	language
0	en	7094
1	de	414
2	af	264
3	it	121
4	fr	112
5	nl	61
6	no	61
7	sv	55
8	es	50
9	zh-cn	50
10	ca	45
11	pl	30
12	da	30
13	pt	19
14	ko	16
15	tl	11
16	et	11
17	ro	10
18	cy	9
19	hr	7
20	sq	6
21	id	5
22	fi	4
23	sl	4
24	so	3
25	vi	2
26	lt	2
27	cs	2
28	lv	1
29	tr	1

x = ticket_df["language"].value_counts()
x=x.sort_index()
plt.figure(figsize=(10,6))
ax= sns.barplot(x=x.index,y= x.values, alpha=0.8)
plt.title("Distribution of text by language")
plt.ylabel('number of records')
plt.xlabel('Language')
rects = ax.patches
labels = x.values
for rect, label in zip(rects, labels):
    height = rect.get_height()
    ax.text(rect.get_x() + rect.get_width()/2, height + 5, label, ha='center', va='bottom')
plt.show();

Observation

• We can see that most of the tickets are in english, followed by tickets in German language.
• We need to translate these into english.

Data Cleaning

ticket_df.head()

	Short description	Description	Caller	Assignment group	Full_Description	language
0	login issue	-verified user details.(employee# & manager na...	spxjnwir pjlcoqds	GRP_0	login issue -verified user details.(employee# ...	en
1	outlook	\r\n\r\nreceived from: hmjdrvpb.komuaywn@gmail...	hmjdrvpb komuaywn	GRP_0	outlook \n\nreceived from: hmjdrvpb.komuaywn@g...	en
2	cant log in to vpn	\r\n\r\nreceived from: eylqgodm.ybqkwiam@gmail...	eylqgodm ybqkwiam	GRP_0	cant log in to vpn \n\nreceived from: eylqgodm...	en
3	unable to access hr_tool page	unable to access hr_tool page	xbkucsvz gcpydteq	GRP_0	unable to access hr_tool page unable to access...	en
4	skype error	skype error	owlgqjme qhcozdfx	GRP_0	skype error skype error	no

Most Common words before data cleaning and removing stop words

wd_counts = Counter()
for i, row in ticket_df.iterrows():
    wd_counts.update(row['Full_Description'].split())
    
wd_counts.most_common(20)

[('to', 8446),
 ('the', 7020),
 ('in', 4909),
 ('is', 3525),
 ('on', 2891),
 ('not', 2851),
 ('and', 2758),
 ('for', 2629),
 ('from:', 2499),
 ('i', 2432),
 ('received', 2371),
 ('a', 2012),
 ('job', 1975),
 ('please', 1946),
 ('password', 1865),
 ('of', 1845),
 ('erp', 1827),
 ('failed', 1700),
 ('job_scheduler', 1629),
 ('at:', 1614)]

### Make text lowercase, remove text in square brackets,remove links,remove punctuation and remove words containing numbers

def clean_text(text):
    '''Make text lowercase, remove text in square brackets,remove links,remove punctuation
    and remove words containing numbers.'''
    text=text.replace(('first name: ').lower(),'firstname')
    text=text.replace(('last name: ').lower(),'lastname')
    text=text.replace(('received from:').lower(),'')
    text=text.replace('email:','')
    text=text.replace('email address:','') 
    index1=text.find('from:')
    index2=text.find('\nsddubject:')
    text=text.replace(text[index1:index2],'')
    index3=text.find('[cid:image')
    index4=text.find(']')
    text=text.replace(text[index3:index4],'')
    text=text.replace('subject:','')
    text=text.replace('received from:','')
    text=text.replace('this message was sent from an unmonitored email address', '')
    text=text.replace('please do not reply to this message', '')
    text=text.replace('monitoring_tool@company.com','MonitoringTool')
    text=text.replace('select the following link to view the disclaimer in an alternate language','')
    text=text.replace('description problem', '') 
    text=text.replace('steps taken far', '')
    text=text.replace('customer job title', '')
    text=text.replace('sales engineer contact', '')
    text=text.replace('description of problem:', '')
    text=text.replace('steps taken so far', '')
    text=text.replace('please do the needful', '')
    text=text.replace('please note that ', '')
    text=text.replace('please find below', '')
    text=text.replace('date and time', '')
    text=text.replace('kindly refer mail', '')
    text=text.replace('name:', '')
    text=text.replace('language:', '')
    text=text.replace('customer number:', '')
    text=text.replace('telephone:', '')
    text=text.replace('summary:', '')
    text=text.replace('sincerely', '')
    text=text.replace('company inc', '')
    text=text.replace('importance:', '')
    text=text.replace('gmail.com', '')
    text=text.replace('company.com', '')
    text=text.replace('microsoftonline.com', '')
    text=text.replace('company.onmicrosoft.com', '')
    text=text.replace('hello', '')
    text=text.replace('hallo', '')
    text=text.replace('hi it team', '')
    text=text.replace('hi team', '')
    text=text.replace('hi', '')
    text=text.replace('best', '')
    text=text.replace('kind', '')
    text=text.replace('regards', '')
    text=text.replace('good morning', '')
    text=text.replace('good afternoon', '')
    text=text.replace('good evening', '')
    text=text.replace('please', '')
    text=text.replace('regards', '')
    text=text.replace('NaN','')
    text=text.replace('can''t','cannot')
    text=text.replace('i''ve','i have')
    text = re.sub(r'\S+@\S+', '', text)
    #custom_punctuation='!"#$%&\'()*+,-./:;<=>?@[\\]^`{|}~'
    text = re.sub(r'\w*\d\w*', '', text)
    text = re.sub(r'\[.*?\]', '', text)
    text = re.sub(r'https?://\S+|www\.\S+', '', text)
    text = re.sub(r'<.*?>+', '', text)
    text = re.sub(r'[%s]' % re.escape(string.punctuation), ' ', text)
    text = re.sub(r'\r\n', '', text)
    text = re.sub(r'\n', '', text)
    text = re.sub(r'\S+@\S+', '', text)
    text = re.sub(r'\t', '', text)

    text = text.lower()
    return text

ticket_df['cleaned_description'] = ticket_df['Full_Description'].apply(lambda x: clean_text(x))

ticket_df.drop(['Full_Description'],axis=1,inplace=True)

ticket_df['cleaned_description'].head()

0    login issue  verified user details  employee  ...
1    outlook  hmjdrvpb komuaywn  team my meetings s...
2    cannot log in to vpn  eylqgodm ybqkwiam i cann...
3    unable to access hr tool page unable to access...
4                            skype error  skype error 
Name: cleaned_description, dtype: object

def replace(str1, str2):
    return str1.replace(str2, '')

ticket_df['cleaned_description'] = ticket_df.apply(lambda row: replace(row['cleaned_description'], row['Caller']), axis=1)

ticket_df['cleaned_description']

0       login issue  verified user details  employee  ...
1       outlook    team my meetings skype meetings etc...
2          cannot log in to vpn   i cannot log on to vpn 
3       unable to access hr tool page unable to access...
4                               skype error  skype error 
                              ...                        
8495    emails not coming in from zz mail    i am not ...
8496    telephony software issue telephony software issue
8497      windows password reset for tifpdchb pedxruyf...
8498    macne não está funcionando i am unable to acce...
8499    an mehreren pc s lassen sich verscedene prgram...
Name: cleaned_description, Length: 8500, dtype: object

from collections import OrderedDict
ticket_df['cleaned_description'] = (ticket_df['cleaned_description'].str.split()
                              .apply(lambda x: OrderedDict.fromkeys(x).keys())
                              .str.join(' '))

ticket_df['cleaned_description'].head()

0    login issue verified user details employee man...
1    outlook team my meetings skype etc are not app...
2                            cannot log in to vpn i on
3                        unable to access hr tool page
4                                          skype error
Name: cleaned_description, dtype: object

wd_counts = Counter()
for i, row in ticket_df.iterrows():
    wd_counts.update(row['cleaned_description'].split())
    
wd_counts.most_common(20)

[('to', 3199),
 ('in', 2752),
 ('the', 2438),
 ('is', 1867),
 ('not', 1670),
 ('on', 1559),
 ('and', 1477),
 ('for', 1431),
 ('at', 1430),
 ('i', 1287),
 ('a', 1117),
 ('it', 1080),
 ('erp', 1053),
 ('password', 1029),
 ('have', 1021),
 ('job', 1007),
 ('of', 996),
 ('scheduler', 968),
 ('tool', 882),
 ('failed', 877)]

uni_df_with_stopw = pd.DataFrame(wd_counts.most_common(20), columns =['Words', 'Count'])

uni_df_with_stopw

	Words	Count
0	to	3199
1	in	2752
2	the	2438
3	is	1867
4	not	1670
5	on	1559
6	and	1477
7	for	1431
8	at	1430
9	i	1287
10	a	1117
11	it	1080
12	erp	1053
13	password	1029
14	have	1021
15	job	1007
16	of	996
17	scheduler	968
18	tool	882
19	failed	877

x = uni_df_with_stopw
x=x.sort_index()
plt.figure(figsize=(10,6))
ax= sns.barplot(x=x.Words, y=x.Count, alpha=0.8)
plt.title("Distribution of text by words")
plt.ylabel('number of records')
plt.xlabel('Language')
plt.xticks(rotation=90)
rects = ax.patches
labels = x.Count
for rect, label in zip(rects, labels):
    height = rect.get_height()
    ax.text(rect.get_x() + rect.get_width()/2, height + 5, label, ha='center', va='bottom')
plt.show();

Need to remove stop words and unnecessary words that do not add meaning to the content.

Remove stop words

stop = stopwords.words('english')
ticket_df['cleaned_description'] = ticket_df['cleaned_description'].apply(lambda x: " ".join(x for x in str(x).split() if x not in stop))
ticket_df['cleaned_description'].head()

0    login issue verified user details employee man...
1    outlook team meetings skype etc appearing cale...
2                                       cannot log vpn
3                           unable access hr tool page
4                                          skype error
Name: cleaned_description, dtype: object

Lemmatization

## Lemmatization
ticket_df['cleaned_description']= ticket_df['cleaned_description'].apply(lambda x: " ".join([Word(word).lemmatize() for word in str(x).split()]))
ticket_df['cleaned_description'].head()

0    login issue verified user detail employee mana...
1    outlook team meeting skype etc appearing calen...
2                                       cannot log vpn
3                           unable access hr tool page
4                                          skype error
Name: cleaned_description, dtype: object

wd_counts = Counter()
for i, row in ticket_df.iterrows():
    wd_counts.update(row['cleaned_description'].split())
    
wd_counts.most_common(20)

[('password', 1176),
 ('erp', 1053),
 ('job', 1029),
 ('scheduler', 968),
 ('issue', 960),
 ('tool', 937),
 ('failed', 877),
 ('unable', 871),
 ('reset', 855),
 ('t', 810),
 ('user', 808),
 ('monitoringtooljob', 785),
 ('access', 679),
 ('error', 673),
 ('account', 660),
 ('need', 659),
 ('working', 608),
 ('company', 604),
 ('help', 559),
 ('email', 550)]

uni_df = pd.DataFrame(wd_counts.most_common(20), columns =['Words', 'Count'])

uni_df

	Words	Count
0	password	1176
1	erp	1053
2	job	1029
3	scheduler	968
4	issue	960
5	tool	937
6	failed	877
7	unable	871
8	reset	855
9	t	810
10	user	808
11	monitoringtooljob	785
12	access	679
13	error	673
14	account	660
15	need	659
16	working	608
17	company	604
18	help	559
19	email	550

x = uni_df
x=x.sort_index()
plt.figure(figsize=(10,6))
ax= sns.barplot(x=x.Words, y=x.Count, alpha=0.8)
plt.title("Distribution of text by words")
plt.ylabel('number of records')
plt.xlabel('Language')
plt.xticks(rotation=90)
rects = ax.patches
labels = x.Count
for rect, label in zip(rects, labels):
    height = rect.get_height()
    ax.text(rect.get_x() + rect.get_width()/2, height + 5, label, ha='center', va='bottom')
plt.show();

Observations

• Before cleaning data and removing stop words the data had common words as articals, consconants, prepositions. etc
• After cleaning the common words now look like words that would help in predicting.

ticket_df['descr_len'] = ticket_df['cleaned_description'].apply(lambda x: len(x))
ticket_df['descr_len'].mean()

86.72976470588236

print(ticket_df['descr_len'].max())
print(ticket_df['descr_len'].min())

3150
0

ticket_df['num_wds'] = ticket_df['cleaned_description'].apply(lambda x: len(x.split()))
ticket_df['num_wds'].mean()

12.17635294117647

ticket_df['num_wds'].describe()

count    8500.000000
mean       12.176353
std        18.091077
min         0.000000
25%         4.000000
50%         7.000000
75%        15.000000
max       428.000000
Name: num_wds, dtype: float64

len(ticket_df[ticket_df['num_wds']==0])

2

Drop tickets with just one word in Description

ticket_df= ticket_df[ticket_df['num_wds']>1]

print(ticket_df['num_wds'].min())

2

def avg_word(sentence):
  words = sentence.split()
  return (sum(len(word) for word in words)/len(words))


ticket_df['avg_word'] = ticket_df['cleaned_description'].apply(lambda x: avg_word(str(x)))
ticket_df.head()

	Short description	Description	Caller	Assignment group	language	cleaned_description	descr_len	num_wds	avg_word
0	login issue	-verified user details.(employee# & manager na...	spxjnwir pjlcoqds	GRP_0	en	login issue verified user detail employee mana...	125	18	6.000000
1	outlook	\r\n\r\nreceived from: hmjdrvpb.komuaywn@gmail...	hmjdrvpb komuaywn	GRP_0	en	outlook team meeting skype etc appearing calen...	75	11	5.909091
2	cant log in to vpn	\r\n\r\nreceived from: eylqgodm.ybqkwiam@gmail...	eylqgodm ybqkwiam	GRP_0	en	cannot log vpn	14	3	4.000000
3	unable to access hr_tool page	unable to access hr_tool page	xbkucsvz gcpydteq	GRP_0	en	unable access hr tool page	26	5	4.400000
4	skype error	skype error	owlgqjme qhcozdfx	GRP_0	no	skype error	11	2	5.000000

ax=ticket_df['num_wds'].plot(kind='hist', bins=50, fontsize=14, figsize=(12,10))
ax.set_title('Description Length in Words\n', fontsize=20)
ax.set_ylabel('Frequency', fontsize=18)
ax.set_xlabel('Number of Words', fontsize=18);

ticket_df['uniq_wds'] = ticket_df['cleaned_description'].str.split().apply(lambda x: len(set(x)))
ticket_df['uniq_wds'].head()

0    18
1    11
2     3
3     5
4     2
Name: uniq_wds, dtype: int64

ticket_df['uniq_wds'].describe()

count    8457.000000
mean       12.107958
std        17.674118
min         2.000000
25%         4.000000
50%         7.000000
75%        15.000000
max       415.000000
Name: uniq_wds, dtype: float64

assign_grps = ticket_df.groupby('Assignment group')
ax=assign_grps['num_wds'].aggregate(np.mean).plot(kind='bar', fontsize=14, figsize=(20,10))
ax.set_title('Mean Number of Words in tickets per Assignment Group\n', fontsize=20)
ax.set_ylabel('Mean Number of Words', fontsize=18)
ax.set_xlabel('Assignment Group', fontsize=18);

ax=assign_grps['uniq_wds'].aggregate(np.mean).plot(kind='bar', fontsize=14, figsize=(20,10))
ax.set_title('Mean Number of Unique Words per tickets in Assignment Group\n', fontsize=20)
ax.set_ylabel('Mean Number of Unique Words', fontsize=18)
ax.set_xlabel('Assignment Group', fontsize=18);

Observations

• Group-56 has the highest amount of words but group-69 has the highest amount of unique words.
• Thus it means the group wth highest words doesnt necessarily have to contain meaningfull content. Also means there are more redundant words in them.

Tokenization

tokenizer = nltk.tokenize.RegexpTokenizer(r'\w+')
ticket_df['tokenized_desc'] = ticket_df['cleaned_description'].apply(lambda x: tokenizer.tokenize(x))

ticket_df['tokenized_desc'].head()

0    [login, issue, verified, user, detail, employe...
1    [outlook, team, meeting, skype, etc, appearing...
2                                   [cannot, log, vpn]
3                     [unable, access, hr, tool, page]
4                                       [skype, error]
Name: tokenized_desc, dtype: object

def combine_text(list_of_text):
    combined_text = ' '.join(list_of_text)
    return combined_text

ticket_df['tokenized_desc'] = ticket_df['tokenized_desc'].apply(lambda x : combine_text(x))

ticket_df.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 8457 entries, 0 to 8499
Data columns (total 11 columns):
 #   Column               Non-Null Count  Dtype  
---  ------               --------------  -----  
 0   Short description    8457 non-null   object 
 1   Description          8457 non-null   object 
 2   Caller               8457 non-null   object 
 3   Assignment group     8457 non-null   object 
 4   language             8457 non-null   object 
 5   cleaned_description  8457 non-null   object 
 6   descr_len            8457 non-null   int64  
 7   num_wds              8457 non-null   int64  
 8   avg_word             8457 non-null   float64
 9   uniq_wds             8457 non-null   int64  
 10  tokenized_desc       8457 non-null   object 
dtypes: float64(1), int64(3), object(7)
memory usage: 792.8+ KB

ticket_df.head()

	Short description	Description	Caller	Assignment group	language	cleaned_description	descr_len	num_wds	avg_word	uniq_wds	tokenized_desc
0	login issue	-verified user details.(employee# & manager na...	spxjnwir pjlcoqds	GRP_0	en	login issue verified user detail employee mana...	125	18	6.000000	18	login issue verified user detail employee mana...
1	outlook	\r\n\r\nreceived from: hmjdrvpb.komuaywn@gmail...	hmjdrvpb komuaywn	GRP_0	en	outlook team meeting skype etc appearing calen...	75	11	5.909091	11	outlook team meeting skype etc appearing calen...
2	cant log in to vpn	\r\n\r\nreceived from: eylqgodm.ybqkwiam@gmail...	eylqgodm ybqkwiam	GRP_0	en	cannot log vpn	14	3	4.000000	3	cannot log vpn
3	unable to access hr_tool page	unable to access hr_tool page	xbkucsvz gcpydteq	GRP_0	en	unable access hr tool page	26	5	4.400000	5	unable access hr tool page
4	skype error	skype error	owlgqjme qhcozdfx	GRP_0	no	skype error	11	2	5.000000	2	skype error

Visualize percentage of tickets per assignment group

# Pie chart
fig1, ax1 = plt.subplots(figsize=(30, 20))
ax1.pie(tkt_df['percentage'], labels=tkt_df['index'], labeldistance=1.1, pctdistance=0.8, rotatelabels=90, autopct='%.2f%%', shadow=True, startangle=90)
# Equal aspect ratio ensures that pie is drawn as a circle
ax1.axis('equal')  
fig1.legend(loc=0, prop={'size': 10})
plt.tight_layout()
plt.show()

Interactive plot

def configure_plotly_browser_state():
  import IPython
  display(IPython.core.display.HTML('''
        <script src="/static/components/requirejs/require.js"></script>
        <script>
          requirejs.config({
            paths: {
              base: '/static/base',
              plotly: 'https://cdn.plot.ly/plotly-1.5.1.min.js?noext',
            },
          });
        </script>
        '''))

configure_plotly_browser_state()

# Assignment group distribution
print('\033[1mTotal assignment groups:\033[0m', ticket_df['Assignment group'].nunique())

# Histogram
# ticket_df['Assignment group'].iplot(kind='hist', title='Assignment Group Distribution- Histogram')

# Pie chart
assgn_grp = pd.DataFrame(ticket_df.groupby('Assignment group').size(),columns = ['Count']).reset_index()
assgn_grp.iplot(kind='pie', labels='Assignment group', values='Count', title='Assignment Group Distribution', hoverinfo="label+percent+name", hole=0.25)

Total assignment groups: 74

Pareto Chart

fig, ax = plt.subplots(1,1,figsize=(30, 15))
bars = ax.bar(tkt_df['index'], tkt_df['Assignment group'], color="C0", label = 'Count')
plt.xticks(rotation=90)
ax2 = ax.twinx()
ax2.plot(tkt_df.index, tkt_df["percentage"], color="C1", marker="D", ms=7, label = 'Percentage')
ax2.yaxis.set_major_formatter(PercentFormatter())

ax.tick_params(axis="y", colors="C0")
ax2.tick_params(axis="y", colors="C1")
lines_1, labels_1 = ax.get_legend_handles_labels()
lines_2, labels_2 = ax2.get_legend_handles_labels()

lines = lines_1 + lines_2
labels = labels_1 + labels_2

ax.legend(lines, labels, loc=0, prop={'size': 20})

ax.set_xlabel("Assignment group")
ax.set_ylabel("Count")
ax2.set_ylabel("Percentage")

plt.title('Assignment groups ticket distribution')

for bar in bars:
    yval = bar.get_height()
    ax.text(bar.get_x(), yval + .005, yval)
plt.show()

Observations

• The data is highly skewed with 46.77% of the tickets present in the Group 0 followed by Group 8, 24, 12, 9, 2, 19
• There are few Groups with less than 10 tickets and few with just 1 or 2 tickets. Have to consider dropping such tickets due to lack of sufficient data for prediction.

Top 20 Assignment groups with highest number of tickets

top_20_df = ticket_df['Assignment group'].value_counts().nlargest(20).reset_index()
top_20_df

	index	Assignment group
0	GRP_0	3956
1	GRP_8	660
2	GRP_24	277
3	GRP_12	257
4	GRP_9	252
5	GRP_2	241
6	GRP_19	215
7	GRP_3	198
8	GRP_6	182
9	GRP_13	145
10	GRP_10	140
11	GRP_5	129
12	GRP_14	118
13	GRP_25	116
14	GRP_33	107
15	GRP_4	100
16	GRP_29	97
17	GRP_18	88
18	GRP_16	85
19	GRP_17	81

plt.figure(figsize=(12,6))
bars = plt.bar(top_20_df['index'],top_20_df['Assignment group'])
plt.title('Top 20 Assignment groups with highest number of Tickets')
plt.xlabel('Assignment Group')
plt.xticks(rotation=90)
plt.ylabel('Number of Tickets')

for bar in bars:
    yval = bar.get_height()
    plt.text(bar.get_x(), yval + .005, yval)
plt.tight_layout()
plt.show()

Bottom 20 Assignment groups with lowest number of tickets

bottom_20_df = ticket_df['Assignment group'].value_counts().nsmallest(20).reset_index()
bottom_20_df

	index	Assignment group
0	GRP_35	1
1	GRP_73	1
2	GRP_64	1
3	GRP_70	1
4	GRP_67	1
5	GRP_61	1
6	GRP_72	2
7	GRP_57	2
8	GRP_69	2
9	GRP_71	2
10	GRP_54	2
11	GRP_58	3
12	GRP_68	3
13	GRP_38	3
14	GRP_63	3
15	GRP_56	3
16	GRP_66	4
17	GRP_32	4
18	GRP_43	5
19	GRP_49	6

plt.figure(figsize=(12,6))
bars = plt.bar(bottom_20_df['index'],bottom_20_df['Assignment group'])
plt.title('Bottom 20 Assignment groups with small number of Tickets')
plt.xlabel('Assignment Group')
plt.xticks(rotation=90)
plt.ylabel('Number of Tickets')
for bar in bars:
    yval = bar.get_height()
    plt.text(bar.get_x(), yval + .005, yval)
plt.tight_layout()
plt.show()

Distribution of tickets within groups less than 30 tickets

# Find out the Assignment Groups with less than equal to 30 tickets assigned
rare_ticket_df = ticket_df.groupby(['Assignment group']).filter(lambda x: len(x) <= 30)
print('Groups with less than equal to 30 tickets assigned:', rare_ticket_df['Assignment group'].nunique())

configure_plotly_browser_state()
rare_ticket_df['Assignment group'].iplot(kind='hist', xTitle='Assignment Group', yTitle='count', colorscale='-orrd', title='Records by rare Assignment Groups- Histogram')

plt.show()

Groups with less than equal to 30 tickets assigned: 40

df_range = pd.DataFrame(columns=['Range','Group_count'])
one_ticket = {'Range':'1 ticket','Group_count':len(tkt_df[tkt_df['Assignment group'] < 2])}
_2_5_ticket = {'Range':'2-5 ticket',
              'Group_count':len(tkt_df[(tkt_df['Assignment group'] > 1)& (tkt_df['Assignment group'] < 6) ])}
_10_ticket = {'Range':' 6-10 ticket',
              'Group_count':len(tkt_df[(tkt_df['Assignment group'] > 5)& (tkt_df['Assignment group'] < 11)])}
_10_20_ticket = {'Range':' 11-20 ticket',
              'Group_count':len(tkt_df[(tkt_df['Assignment group'] > 10)& (tkt_df['Assignment group'] < 21)])}
_20_50_ticket = {'Range':' 21-50 ticket',
              'Group_count':len(tkt_df[(tkt_df['Assignment group'] > 20)& (tkt_df['Assignment group'] < 51)])}
_51_100_ticket = {'Range':' 51-100 ticket',
              'Group_count':len(tkt_df[(tkt_df['Assignment group'] > 50)& (tkt_df['Assignment group'] < 101)])}
_100_ticket = {'Range':' >100 ticket',
              'Group_count':len(tkt_df[(tkt_df['Assignment group'] > 100)])}
#append row to the dataframe
df_range = df_range.append([one_ticket,_2_5_ticket,_10_ticket,
                          _10_20_ticket,_20_50_ticket,_51_100_ticket,_100_ticket], ignore_index=True)

df_range

	Range	Group_count
0	1 ticket	6
1	2-5 ticket	13
2	6-10 ticket	6
3	11-20 ticket	9
4	21-50 ticket	16
5	51-100 ticket	9
6	>100 ticket	15

plt.figure(figsize=(10, 8))
plt.pie(df_range['Group_count'],labels=df_range['Range'],autopct='%1.1f%%', startangle=15, shadow = True);
plt.title('Assignment Groups Distribution')
plt.axis('equal');

Observations

• Top 5 groups with hightest tickets are - 0, 8, 24, 12, 9
• Bottom 5 groups with lowest tickets are - 70, 64, 67, 73, 35
• There are 40 groups with just 30 or less tickets assigned, amongst which 6 groups happened to be assigned with just 1 ticket, 4 groups with just 2 tickets each and 5 groups with 3 tickets each.
• There are 15 assignment groups that has more than 100 tickets which accounts to only 20.3% of the overall dataset.

ticket_df.head()

	Short description	Description	Caller	Assignment group	language	cleaned_description	descr_len	num_wds	avg_word	uniq_wds	tokenized_desc
0	login issue	-verified user details.(employee# & manager na...	spxjnwir pjlcoqds	GRP_0	en	login issue verified user detail employee mana...	125	18	6.000000	18	login issue verified user detail employee mana...
1	outlook	\r\n\r\nreceived from: hmjdrvpb.komuaywn@gmail...	hmjdrvpb komuaywn	GRP_0	en	outlook team meeting skype etc appearing calen...	75	11	5.909091	11	outlook team meeting skype etc appearing calen...
2	cant log in to vpn	\r\n\r\nreceived from: eylqgodm.ybqkwiam@gmail...	eylqgodm ybqkwiam	GRP_0	en	cannot log vpn	14	3	4.000000	3	cannot log vpn
3	unable to access hr_tool page	unable to access hr_tool page	xbkucsvz gcpydteq	GRP_0	en	unable access hr tool page	26	5	4.400000	5	unable access hr tool page
4	skype error	skype error	owlgqjme qhcozdfx	GRP_0	no	skype error	11	2	5.000000	2	skype error

Drop groups that has just one ticket each

ticket_df.drop(ticket_df[(ticket_df['Assignment group'] == 'GRP_73')].index , inplace=True)
ticket_df.drop(ticket_df[(ticket_df['Assignment group'] == 'GRP_60')].index , inplace=True)
ticket_df.drop(ticket_df[(ticket_df['Assignment group'] == 'GRP_35')].index , inplace=True)
ticket_df.drop(ticket_df[(ticket_df['Assignment group'] == 'GRP_70')].index , inplace=True)
ticket_df.drop(ticket_df[(ticket_df['Assignment group'] == 'GRP_64')].index , inplace=True)
ticket_df.drop(ticket_df[(ticket_df['Assignment group'] == 'GRP_67')].index , inplace=True)

ticket_df.shape

(8432, 11)

ticket_df.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 8432 entries, 0 to 8499
Data columns (total 11 columns):
 #   Column               Non-Null Count  Dtype  
---  ------               --------------  -----  
 0   Short description    8432 non-null   object 
 1   Description          8432 non-null   object 
 2   Caller               8432 non-null   object 
 3   Assignment group     8432 non-null   object 
 4   language             8432 non-null   object 
 5   cleaned_description  8432 non-null   object 
 6   descr_len            8432 non-null   int64  
 7   num_wds              8432 non-null   int64  
 8   avg_word             8432 non-null   float64
 9   uniq_wds             8432 non-null   int64  
 10  tokenized_desc       8432 non-null   object 
dtypes: float64(1), int64(3), object(7)
memory usage: 790.5+ KB

Word Cloud

def f_word_cloud(column):
    
    comment_words = ' '
    stopwords = set(STOPWORDS)

    # iterate through the csv file 
    for val in column: 

        # typecaste each val to string 
        val = str(val) 

        # split the value 
        tokens = val.split() 

        # Converts each token into lowercase 
        for i in range(len(tokens)): 
            tokens[i] = tokens[i].lower() 

        for words in tokens: 
            comment_words = comment_words + words + ' '


    wordcloud = WordCloud(width = 800, height = 800, 
                    background_color ='white', 
                    stopwords = stopwords, 
                    min_font_size = 10).generate(comment_words) 
    
    return wordcloud

wordcloud = f_word_cloud(ticket_df.cleaned_description)
# plot the WordCloud image                        
plt.figure(figsize = (8, 8), facecolor = None) 
plt.imshow(wordcloud) 
plt.axis("off") 
plt.tight_layout(pad = 0) 
  
plt.show()

Top 4 Group word cloud to understand the group's field of operation

Group 0

wordcloud = f_word_cloud(ticket_df[ticket_df['Assignment group']=='GRP_0'].cleaned_description)
# plot the WordCloud image                        
plt.figure(figsize = (8, 8), facecolor = None) 
plt.imshow(wordcloud) 
plt.axis("off") 
plt.tight_layout(pad = 0) 
  
plt.show()

ticket_df[ticket_df["cleaned_description"].str.contains("account lock")]["Assignment group"].value_counts()

GRP_0     242
GRP_2       5
GRP_1       1
GRP_72      1
GRP_34      1
GRP_31      1
GRP_14      1
Name: Assignment group, dtype: int64

ticket_df[ticket_df["cleaned_description"].str.contains("password reset")]["Assignment group"].value_counts()

GRP_0     343
GRP_2       5
GRP_7       2
GRP_12      1
Name: Assignment group, dtype: int64

Observations

• Group 0 seems to have ticket issue related to password resert, access issues , login issue, etc. Seem to be an L1 category issues to be resolved.
• Groups that have similar common words as group-0 are group- 2,7,12,14,17 etc.

wordcloud = f_word_cloud(ticket_df[ticket_df['Assignment group']=='GRP_8'].cleaned_description)
# plot the WordCloud image                        
plt.figure(figsize = (8, 8), facecolor = None) 
plt.imshow(wordcloud) 
plt.axis("off") 
plt.tight_layout(pad = 0) 
  
plt.show()

ticket_df[ticket_df["cleaned_description"].str.contains("jobscheduler")]["Assignment group"].value_counts()

Series([], Name: Assignment group, dtype: int64)

ticket_df[ticket_df["cleaned_description"].str.contains("outage")]["Assignment group"].value_counts()

GRP_8     141
GRP_4      17
GRP_0      10
GRP_37      1
GRP_16      1
Name: Assignment group, dtype: int64

Observations

• Group 8 have issues related to monitoring tool, scheduled maintenance, outage, job failure, etc.
• These issues seems to be of higher priority and thus it cam be assumed to be L3 related issue.
• Groups that has similar ticket issue key words as that of group 8 are group-9,6,5,10,4,47,0,45,12,1,13,14,29,18 etc.

We find that the similar groups in group 0 and 8 do not overlap and can see clear distinction in the type of field and level of work

Comparison between Group-0(L1/L2) and Group-8(L3)

ticket_df_level = ticket_df.copy()
ticket_df_level['Target'] = np.where(ticket_df_level['Assignment group']=='GRP_0','L1/L2',np.where(ticket_df_level['Assignment group'] =='GRP_8','L1/L2','L3'))

x=ticket_df_level.Target.value_counts()
sns.barplot(x.index,x)
plt.gca().set_ylabel('tickets')

Text(0, 0.5, 'tickets')

Observations - L1/L2 type ticket counts are quiet higher than L3 tickets in the dataset

Group-12

wordcloud = f_word_cloud(ticket_df[ticket_df['Assignment group']=='GRP_12'].cleaned_description)
# plot the WordCloud image                        
plt.figure(figsize = (8, 8), facecolor = None) 
plt.imshow(wordcloud) 
plt.axis("off") 
plt.tight_layout(pad = 0) 
  
plt.show()

Observations - Group-12 tickets mostly revolve around server, asa deny, dst outside, outside access.

Group-24

wordcloud = f_word_cloud(ticket_df[ticket_df['Assignment group']=='GRP_24'].cleaned_description)
# plot the WordCloud image                        
plt.figure(figsize = (8, 8), facecolor = None) 
plt.imshow(wordcloud) 
plt.axis("off") 
plt.tight_layout(pad = 0) 
  
plt.show()

Observations - Group-24 consist of german language which needs translation

Caller Distribution Analysis

len(ticket_df['Caller'].unique())

2942

ticket_df.groupby("Caller").size()

Caller
DEsowkrH sciejzNE      2
HdOuWxAt qfKdlJMx      1
ZkBogxib QsEJzdZO    151
abiuclvj brziktno      1
abtknpew ifcekjdb      1
                    ... 
zylaexnv pvkzbduh      3
zylwdbig wdkbztjp      1
zymdwqsi jzvbthil      1
zywoxerf paqxtrfk      9
zyxjagro vjgozhpn      2
Length: 2942, dtype: int64

wordcloud = f_word_cloud(ticket_df.Caller)
# plot the WordCloud image                        
plt.figure(figsize = (8, 8), facecolor = None) 
plt.imshow(wordcloud) 
plt.axis("off") 
plt.tight_layout(pad = 0) 
  
plt.show()

fig1 = px.histogram(ticket_df, x="Caller")
HTML(fig1.to_html())

top_10_caller_df = ticket_df['Caller'].value_counts().nlargest(10).reset_index()
top_10_caller_df

	index	Caller
0	bpctwhsn kzqsbmtp	794
1	ZkBogxib QsEJzdZO	151
2	fumkcsji sarmtlhy	134
3	rbozivdq gmlhrtvp	79
4	rkupnshb gsmzfojw	71
5	jloygrwh acvztedi	62
6	spxqmiry zpwgoqju	62
7	oldrctiu bxurpsyi	57
8	olckhmvx pcqobjnd	54
9	dkmcfreg anwmfvlg	51

plt.figure(figsize=(12,6))
bars = plt.bar(top_10_caller_df['index'],top_10_caller_df['Caller'])
plt.title('Top 20 callers with highest number of tickets')
plt.xlabel('Caller')
plt.xticks(rotation=90)
plt.ylabel('Number of Tickets')

for bar in bars:
    yval = bar.get_height()
    plt.text(bar.get_x(), yval + .005, yval)
plt.tight_layout()
plt.show()

Top 4 caller word cloud

wordcloud = f_word_cloud(ticket_df[ticket_df['Caller']=='bpctwhsn kzqsbmtp'].cleaned_description)
# plot the WordCloud image                        
plt.figure(figsize = (8, 8), facecolor = None) 
plt.imshow(wordcloud) 
plt.axis("off") 
plt.tight_layout(pad = 0) 
  
plt.show()

wordcloud = f_word_cloud(ticket_df[ticket_df['Caller']=='ZkBogxib QsEJzdZO'].cleaned_description)
# plot the WordCloud image                        
plt.figure(figsize = (8, 8), facecolor = None) 
plt.imshow(wordcloud) 
plt.axis("off") 
plt.tight_layout(pad = 0) 
  
plt.show()

wordcloud = f_word_cloud(ticket_df[ticket_df['Caller']=='fumkcsji sarmtlhy'].cleaned_description)
# plot the WordCloud image                        
plt.figure(figsize = (8, 8), facecolor = None) 
plt.imshow(wordcloud) 
plt.axis("off") 
plt.tight_layout(pad = 0) 
  
plt.show()

wordcloud = f_word_cloud(ticket_df[ticket_df['Caller']=='rbozivdq gmlhrtvp'].cleaned_description)
# plot the WordCloud image                        
plt.figure(figsize = (8, 8), facecolor = None) 
plt.imshow(wordcloud) 
plt.axis("off") 
plt.tight_layout(pad = 0) 
  
plt.show()

Observations

• The top caller "bpctwhsn kzqsbmtp" & "ZkBogxib QsEJzdZO" mostly addresses issues related to job_scheduler, monitoring tool.
• Caller "fumkcsji sarmtlhy" tickets are related to ticket updates and specific ticket related issues.
• Caller "rbozivdq gmlhrtvp" are mostly related to basic front desk resolvable issues such as loud noise, general enquiry, loud noise etc.

# Top 5 callers in each assignment group
top_n = 5
s = ticket_df['Caller'].groupby(ticket_df['Assignment group']).value_counts()
caller_grp = pd.DataFrame(s.groupby(level=0).nlargest(top_n).reset_index(level=0, drop=True))
caller_grp.head(15)

		Caller
Assignment group	Caller
GRP_0	fumkcsji sarmtlhy	132
	rbozivdq gmlhrtvp	78
	olckhmvx pcqobjnd	54
	efbwiadp dicafxhv	45
	mfeyouli ndobtzpw	13
GRP_1	bpctwhsn kzqsbmtp	6
	jloygrwh acvztedi	4
	jyoqwxhz clhxsoqy	3
	spxqmiry zpwgoqju	3
	kbnfxpsy gehxzayq	2
GRP_10	bpctwhsn kzqsbmtp	60
	ihfkwzjd erbxoyqk	6
	dizquolf hlykecxa	5
	gnasmtvx cwxtsvkm	3
	hlrmufzx qcdzierm	3

# Visualize Top 5 callers in each of top 10 assignment groups
top_n = 10 
top_grps = assgn_grp.nlargest(top_n, 'Count')['Assignment group'].tolist()

fig_cols = 5
fig_rows = int(np.ceil(top_n/fig_cols))
fig, axes = plt.subplots(fig_rows, fig_cols, figsize=(13,9.5))
fig.suptitle('Top 5 callers in each of top 10 assignment groups- Pie Chart', y=1, va= 'bottom', size='20')
for row in range(fig_rows):
    for col in range(fig_cols):
        grp_n = fig_cols * row + col
        if grp_n < top_n:
            xs = caller_grp.xs(top_grps[grp_n])
            _ = axes[row,col].pie(xs, autopct='%1.1f%%', explode=[0.05]*5)
            axes[row,col].legend(labels=xs.index,loc="best")
            axes[row,col].axis('equal')
            axes[row,col].set_title(top_grps[grp_n])

plt.tight_layout()

# Check if any caller appears to raise ticket for multiple groups
mul_caller = caller_grp[caller_grp.Caller.duplicated()]
uni_mul_caller = [idx[1] for idx in mul_caller.index[mul_caller.Caller.unique()]]
print(f'\033[1mFollowing {len(uni_mul_caller)} callers happen to raise tickets for multiple groups:\033[0m\n')
print(uni_mul_caller)

mul_caller

Following 16 callers happen to raise tickets for multiple groups:

['hlrmufzx qcdzierm', 'fbgetczn jlsvxura', 'gnasmtvx cwxtsvkm', 'ihfkwzjd erbxoyqk', 'tqfnalpj qyoscnge', 'fmqubnvs kcxpeyiv', 'tghrloks jbgcvlmf', 'jwqyxbzs adpvilqu', 'nuhfwplj ojcwxser', 'oldrctiu bxurpsyi', 'vlymsnej whlqxcst', 'dkmcfreg anwmfvlg', 'bpctwhsn kzqsbmtp', 'lmsxcvoz vzhkdpfn', 'spxqmiry zpwgoqju', 'obanjrhg rnafleys']

		Caller
Assignment group	Caller
GRP_1	spxqmiry zpwgoqju	3
GRP_10	ihfkwzjd erbxoyqk	6
	gnasmtvx cwxtsvkm	3
	hlrmufzx qcdzierm	3
GRP_11	tghrloks jbgcvlmf	2
...	...	...
GRP_72	ydigzqbu xdgjizek	1
GRP_8	ZkBogxib QsEJzdZO	54
GRP_9	ctzykflo evzbhgru	3
	sholvcmf bjtpomrl	3
	urhpnlaf agmsfqil	3

273 rows × 1 columns

Observations

• There are overall 2950 callers in the dataset.
• 15 callers are involved in raising tickets for multiple assignment group, overall contributing to 281 tickets in the dataset.
• The top 5 caller ticket distribution in every group is widely distributed almost having a balanced distribution. This indicated that most of the group resolves issues widely over various field impacting business

Data Pre-Processing

N-Grams

Uni-Gram

def get_top_n_words(corpus, range=(1,1), n=None):
    vec = CountVectorizer(ngram_range=range, stop_words = 'english').fit(corpus)
    bag_of_words = vec.transform(corpus)
    sum_words = bag_of_words.sum(axis=0) 
    words_freq = [(word, sum_words[0, idx]) for word, idx in vec.vocabulary_.items()]
    words_freq =sorted(words_freq, key = lambda x: x[1], reverse=True)
    return words_freq[:n]


common_words = get_top_n_words(ticket_df['cleaned_description'], (1,1),20)
for word, freq in common_words:
    print(word, freq)

password 1176
erp 1052
job 1016
issue 957
scheduler 955
tool 933
unable 868
failed 863
reset 855
user 808
monitoringtooljob 772
access 677
error 673
account 660
need 658
working 607
company 602
help 558
email 548
ticket 528

Bi-Grams

# The distribution of top bigrams after removing stop words for Complete_Description

common_words = get_top_n_words(ticket_df['cleaned_description'], (2,2),20)
for word, freq in common_words:
    print(word, freq)

failed scheduler 796
scheduler monitoringtooljob 768
job failed 734
password reset 351
reset password 241
engineering tool 212
management tool 209
account locked 208
collaboration platform 181
microsoft internet 166
password management 165
job scheduler 159
internet explorer 156
error message 156
scheduler monitoringtool 155
ticket update 153
abended job 151
phone email 147
specify cert 143
scheduled maintenance 143

Tri-Grams

# The distribution of top trigrams after removing stop words for Complete_Description

common_words = get_top_n_words(ticket_df['cleaned_description'], (3,3) ,20)
for word, freq in common_words:
    print(word, freq)

failed scheduler monitoringtooljob 768
job failed scheduler 733
password management tool 159
abended job scheduler 151
contact notified phone 143
notified phone email 143
remote dial equipment 143
dial equipment reset 143
equipment reset verified 143
reset verified working 143
verified working vendor 142
working vendor verizon 142
gsc started additional 142
started additional diagnostics 142
start scheduled maintenance 141
provider maint ticket 141
phone email remote 141
email remote dial 141
verizon gsc started 141
acti contact notified 140

Feature Extraction

Calculating TF-IDF

# Total no. of unique words in "tokenized_desc" column
len(Counter(" ".join(ticket_df['tokenized_desc'].str.lower().values.tolist()).split(" ")).items())

12270

tfidf = TfidfVectorizer(max_features=12270, analyzer = 'word', min_df=2, max_df=0.95, ngram_range=(1, 2))
tckt_tfidf = tfidf.fit_transform(ticket_df['tokenized_desc'])

len(tckt_tfidf.todense())

8432

#  create a dictionary mapping the tokens to their tfidf values
tfidf = dict(zip(tfidf.get_feature_names(), tfidf.idf_))
tfidf = pd.DataFrame(columns=['tfidf']).from_dict(dict(tfidf), orient='index')
tfidf.columns = ['tfidf']

tfidf.sort_values(by=['tfidf'], ascending=True).head(10)

	tfidf
erp	3.080509
password	3.085269
job	3.120224
scheduler	3.177150
issue	3.207954
tool	3.226463
unable	3.272565
failed	3.278335
reset	3.288806
failed scheduler	3.359053

tfidf.sort_values(by=['tfidf'], ascending=False).head(20)

	tfidf
list attached	8.941296
srinfhyathbrowser microsoft	8.941296
germany site	8.941296
germany registergerirtcht	8.941296
germany plant	8.941296
germany pbx	8.941296
germany office	8.941296
germany new	8.941296
germany log	8.941296
srv microsoft	8.941296
germany funktioniert	8.941296
germany erp	8.941296
germany dmvpn	8.941296
srv warning	8.941296
ssl certificate	8.941296
german time	8.941296
german english	8.941296
st et	8.941296
gergryth	8.941296
gerade	8.941296

plt.figure(figsize=(15,7))
sns.distplot(tfidf["tfidf"])

<matplotlib.axes._subplots.AxesSubplot at 0x7f6048cd3150>

Observations

• Low TF-IDF - Words that have low significance in the dataset - Erp, job, passwords, jobscheduler, failed etc.
• High TF-IDF - Words that have high significance in the dataset - outside accessgroup, dst outside, src inside, src, accessgroup etc.
• TF-IDF distribution curve suggest that it has more dimensions to it. Thus it would be apt to apply some dimensionality reduction techniques like LDA.

Perform LDA - Dimensionality Reduction

# create count vectorizer first
cvectorizer = CountVectorizer(min_df=4, max_features=637, ngram_range=(1,2))
cvz = cvectorizer.fit_transform(ticket_df['tokenized_desc'])

# generate topic models using Latent Dirichlet Allocation
lda_model = LatentDirichletAllocation(n_components=10, learning_method='online', max_iter=20, random_state=42)
X_topics = lda_model.fit_transform(cvz)

n_top_words = 10
topic_summaries = []

# get topics and topic terms
topic_word = lda_model.components_ 
vocab = cvectorizer.get_feature_names()

for i, topic_dist in enumerate(topic_word):
    topic_words = np.array(vocab)[np.argsort(topic_dist)][:-(n_top_words+1):-1]
    topic_summaries.append(' '.join(topic_words))
    print('Topic {}: {}'.format(i, ' | '.join(topic_words)))

Topic 0: password | reset | password reset | erp | microsoft | skype | internet | microsoft internet | explorer | request
Topic 1: job | scheduler | failed | failed scheduler | monitoringtooljob | scheduler monitoringtooljob | job failed | monitoringtool | job scheduler | abended
Topic 2: password | update | tool | management | ticket | management tool | manager | using | password management | reset
Topic 3: need | issue | message | help | access | system | erp | new | error | printer
Topic 4: pc | mit | nicht | laptop | bitte | germany | für | probleme | und | der
Topic 5: et | since | company | acti | start | na | power | contact | outage | backup
Topic 6: call | work | time | screen | connection | user | know | event | software | let
Topic 7: outlook | user | issue | access | id | unable | open | email | crm | working
Topic 8: tool | server | problem | error | engineering | see | engineering tool | erp | log | attached
Topic 9: account | unable | locked | erp | account locked | unlock | login | check | connect | unable login

Observations

• We can see that password related tickets are classified as - topic 5, account related tickets in - topic 2 , job scheduler related tickets in - topic 9 etc.
• Thus each topic revolves around a specific ticket issue.

Tranforming tokens to vector using TF-IDF

# collect the tfid matrix in numpy array
array = tckt_tfidf.todense()

# store the tf-idf array into pandas dataframe
df_tfidf = pd.DataFrame(array)
df_tfidf.head(10)

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	12230	12231	12232	12233	12234	12235	12236	12237	12238	12239	12240	12241	12242	12243	12244	12245	12246	12247	12248	12249	12250	12251	12252	12253	12254	12255	12256	12257	12258	12259	12260	12261	12262	12263	12264	12265	12266	12267	12268	12269
0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.124129	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
1	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
2	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
3	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
4	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
5	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
6	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
7	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
8	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
9	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0

10 rows × 12270 columns

#Check for equality in no of rows before and presence of null values before copying word count columns into TF-IDF vectorized dataframe
print(ticket_df[ticket_df.isnull().values == True])
print(ticket_df.shape)
print(df_tfidf.shape)

Empty DataFrame
Columns: [Short description, Description, Caller, Assignment group, language, cleaned_description, descr_len, num_wds, avg_word, uniq_wds, tokenized_desc]
Index: []
(8432, 11)
(8432, 12270)

df_tfidf['descr_len']= ticket_df['descr_len'].values
df_tfidf['num_wds']= ticket_df['num_wds'].values
df_tfidf['avg_word']= ticket_df['avg_word'].values
df_tfidf['uniq_wds']= ticket_df['uniq_wds'].values
df_tfidf.head()

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	12234	12235	12236	12237	12238	12239	12240	12241	12242	12243	12244	12245	12246	12247	12248	12249	12250	12251	12252	12253	12254	12255	12256	12257	12258	12259	12260	12261	12262	12263	12264	12265	12266	12267	12268	12269	descr_len	num_wds	avg_word	uniq_wds
0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.124129	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	125	18	6.000000	18
1	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	75	11	5.909091	11
2	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	14	3	4.000000	3
3	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	26	5	4.400000	5
4	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.000000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	11	2	5.000000	2

5 rows × 12274 columns

# Apply Z-score to normalize the values
df_tfidf_z = df_tfidf.apply(zscore)  # convert all attributes to Z scale 
df_tfidf_z.head()

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	12234	12235	12236	12237	12238	12239	12240	12241	12242	12243	12244	12245	12246	12247	12248	12249	12250	12251	12252	12253	12254	12255	12256	12257	12258	12259	12260	12261	12262	12263	12264	12265	12266	12267	12268	12269	descr_len	num_wds	avg_word	uniq_wds
0	-0.025421	-0.026586	-0.018833	-0.015403	-0.015244	-0.017794	-0.015398	-0.028546	-0.026633	-0.134593	-0.133797	-0.035485	-0.014615	-0.021332	-0.028877	-0.018748	-0.015375	-0.015375	3.517013	-0.052115	-0.031898	-0.015403	-0.014269	-0.015355	-0.038902	-0.024306	-0.020258	-0.020395	-0.029478	-0.014974	-0.01388	-0.015373	-0.06417	-0.015273	-0.0399	-0.060723	-0.015301	-0.015352	-0.040026	-0.014754	...	-0.021758	-0.021758	-0.015244	-0.015388	-0.021758	-0.021758	-0.015403	-0.015403	-0.01883	-0.015403	-0.01883	-0.01883	-0.01883	-0.01883	-0.015293	-0.01883	-0.01883	-0.033862	-0.015339	-0.015403	-0.015403	-0.015403	-0.021758	-0.026338	-0.015403	-0.025995	-0.018812	-0.01883	-0.01883	-0.01883	-0.01883	-0.021591	-0.01533	-0.015361	-0.021758	-0.021758	0.276837	0.317553	-0.240602	0.332661
1	-0.025421	-0.026586	-0.018833	-0.015403	-0.015244	-0.017794	-0.015398	-0.028546	-0.026633	-0.134593	-0.133797	-0.035485	-0.014615	-0.021332	-0.028877	-0.018748	-0.015375	-0.015375	-0.199857	-0.052115	-0.031898	-0.015403	-0.014269	-0.015355	-0.038902	-0.024306	-0.020258	-0.020395	-0.029478	-0.014974	-0.01388	-0.015373	-0.06417	-0.015273	-0.0399	-0.060723	-0.015301	-0.015352	-0.040026	-0.014754	...	-0.021758	-0.021758	-0.015244	-0.015388	-0.021758	-0.021758	-0.015403	-0.015403	-0.01883	-0.015403	-0.01883	-0.01883	-0.01883	-0.01883	-0.015293	-0.01883	-0.01883	-0.033862	-0.015339	-0.015403	-0.015403	-0.015403	-0.021758	-0.026338	-0.015403	-0.025995	-0.018812	-0.01883	-0.01883	-0.01883	-0.01883	-0.021591	-0.01533	-0.015361	-0.021758	-0.021758	-0.089113	-0.068373	-0.308770	-0.062990
2	-0.025421	-0.026586	-0.018833	-0.015403	-0.015244	-0.017794	-0.015398	-0.028546	-0.026633	-0.134593	-0.133797	-0.035485	-0.014615	-0.021332	-0.028877	-0.018748	-0.015375	-0.015375	-0.199857	-0.052115	-0.031898	-0.015403	-0.014269	-0.015355	-0.038902	-0.024306	-0.020258	-0.020395	-0.029478	-0.014974	-0.01388	-0.015373	-0.06417	-0.015273	-0.0399	-0.060723	-0.015301	-0.015352	-0.040026	-0.014754	...	-0.021758	-0.021758	-0.015244	-0.015388	-0.021758	-0.021758	-0.015403	-0.015403	-0.01883	-0.015403	-0.01883	-0.01883	-0.01883	-0.01883	-0.015293	-0.01883	-0.01883	-0.033862	-0.015339	-0.015403	-0.015403	-0.015403	-0.021758	-0.026338	-0.015403	-0.025995	-0.018812	-0.01883	-0.01883	-0.01883	-0.01883	-0.021591	-0.01533	-0.015361	-0.021758	-0.021758	-0.535572	-0.509431	-1.740291	-0.515163
3	-0.025421	-0.026586	-0.018833	-0.015403	-0.015244	-0.017794	-0.015398	-0.028546	-0.026633	-0.134593	-0.133797	-0.035485	-0.014615	-0.021332	-0.028877	-0.018748	-0.015375	-0.015375	-0.199857	-0.052115	-0.031898	-0.015403	-0.014269	-0.015355	-0.038902	-0.024306	-0.020258	-0.020395	-0.029478	-0.014974	-0.01388	-0.015373	-0.06417	-0.015273	-0.0399	-0.060723	-0.015301	-0.015352	-0.040026	-0.014754	...	-0.021758	-0.021758	-0.015244	-0.015388	-0.021758	-0.021758	-0.015403	-0.015403	-0.01883	-0.015403	-0.01883	-0.01883	-0.01883	-0.01883	-0.015293	-0.01883	-0.01883	-0.033862	-0.015339	-0.015403	-0.015403	-0.015403	-0.021758	-0.026338	-0.015403	-0.025995	-0.018812	-0.01883	-0.01883	-0.01883	-0.01883	-0.021591	-0.01533	-0.015361	-0.021758	-0.021758	-0.447744	-0.399166	-1.440353	-0.402120
4	-0.025421	-0.026586	-0.018833	-0.015403	-0.015244	-0.017794	-0.015398	-0.028546	-0.026633	-0.134593	-0.133797	-0.035485	-0.014615	-0.021332	-0.028877	-0.018748	-0.015375	-0.015375	-0.199857	-0.052115	-0.031898	-0.015403	-0.014269	-0.015355	-0.038902	-0.024306	-0.020258	-0.020395	-0.029478	-0.014974	-0.01388	-0.015373	-0.06417	-0.015273	-0.0399	-0.060723	-0.015301	-0.015352	-0.040026	-0.014754	...	-0.021758	-0.021758	-0.015244	-0.015388	-0.021758	-0.021758	-0.015403	-0.015403	-0.01883	-0.015403	-0.01883	-0.01883	-0.01883	-0.01883	-0.015293	-0.01883	-0.01883	-0.033862	-0.015339	-0.015403	-0.015403	-0.015403	-0.021758	-0.026338	-0.015403	-0.025995	-0.018812	-0.01883	-0.01883	-0.01883	-0.01883	-0.021591	-0.01533	-0.015361	-0.021758	-0.021758	-0.557529	-0.564563	-0.990446	-0.571684

5 rows × 12274 columns

df_tfidf_z['Assignment_group']= ticket_df['Assignment group'].values
df_tfidf_z[df_tfidf_z.isnull().values == True]

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	12235	12236	12237	12238	12239	12240	12241	12242	12243	12244	12245	12246	12247	12248	12249	12250	12251	12252	12253	12254	12255	12256	12257	12258	12259	12260	12261	12262	12263	12264	12265	12266	12267	12268	12269	descr_len	num_wds	avg_word	uniq_wds	Assignment_group

0 rows × 12275 columns

Save the cleansed vectorized and non vectorized dataset into a csv file for future modelling purpose

df_tfidf_z.to_csv('/content/drive/My Drive/Colab Notebooks/Capstone_Project_AIML/tfidf_ticket_df.csv', index=False, encoding='utf_8_sig')
ticket_df.to_csv('/content/drive/My Drive/Colab Notebooks/Capstone_Project_AIML/cleaned_ticket_df.csv', index=False, encoding='utf_8_sig')

Label Encoding / One Hot Encoding

ticket_df = pd.read_csv('/content/drive/My Drive/Colab Notebooks/Capstone_Project_AIML/cleaned_ticket_df.csv',encoding='utf-8')
df_tfidf_z = pd.read_csv('/content/drive/My Drive/Colab Notebooks/Capstone_Project_AIML/tfidf_ticket_df.csv',encoding='utf-8')

Label Encoding

#Label encoding vectorized dataframe
le1 = preprocessing.LabelEncoder()
df_tfidf_z['Assignment_group'] = df_tfidf_z['Assignment_group'].astype(str)
df_tfidf_z["LabelEncodings"] = le1.fit_transform(df_tfidf_z['Assignment_group'])
y_classes_len = len(le1.classes_)
le1.classes_
print("No. of unique target group", y_classes_len)
df_tfidf_z.head()

No. of unique target group 68

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39	...	12236	12237	12238	12239	12240	12241	12242	12243	12244	12245	12246	12247	12248	12249	12250	12251	12252	12253	12254	12255	12256	12257	12258	12259	12260	12261	12262	12263	12264	12265	12266	12267	12268	12269	descr_len	num_wds	avg_word	uniq_wds	Assignment_group	LabelEncodings
0	-0.025421	-0.026586	-0.018833	-0.015403	-0.015244	-0.017794	-0.015398	-0.028546	-0.026633	-0.134593	-0.133797	-0.035485	-0.014615	-0.021332	-0.028877	-0.018748	-0.015375	-0.015375	3.517013	-0.052115	-0.031898	-0.015403	-0.014269	-0.015355	-0.038902	-0.024306	-0.020258	-0.020395	-0.029478	-0.014974	-0.01388	-0.015373	-0.06417	-0.015273	-0.0399	-0.060723	-0.015301	-0.015352	-0.040026	-0.014754	...	-0.015244	-0.015388	-0.021758	-0.021758	-0.015403	-0.015403	-0.01883	-0.015403	-0.01883	-0.01883	-0.01883	-0.01883	-0.015293	-0.01883	-0.01883	-0.033862	-0.015339	-0.015403	-0.015403	-0.015403	-0.021758	-0.026338	-0.015403	-0.025995	-0.018812	-0.01883	-0.01883	-0.01883	-0.01883	-0.021591	-0.01533	-0.015361	-0.021758	-0.021758	0.276837	0.317553	-0.240602	0.332661	GRP_0	0
1	-0.025421	-0.026586	-0.018833	-0.015403	-0.015244	-0.017794	-0.015398	-0.028546	-0.026633	-0.134593	-0.133797	-0.035485	-0.014615	-0.021332	-0.028877	-0.018748	-0.015375	-0.015375	-0.199857	-0.052115	-0.031898	-0.015403	-0.014269	-0.015355	-0.038902	-0.024306	-0.020258	-0.020395	-0.029478	-0.014974	-0.01388	-0.015373	-0.06417	-0.015273	-0.0399	-0.060723	-0.015301	-0.015352	-0.040026	-0.014754	...	-0.015244	-0.015388	-0.021758	-0.021758	-0.015403	-0.015403	-0.01883	-0.015403	-0.01883	-0.01883	-0.01883	-0.01883	-0.015293	-0.01883	-0.01883	-0.033862	-0.015339	-0.015403	-0.015403	-0.015403	-0.021758	-0.026338	-0.015403	-0.025995	-0.018812	-0.01883	-0.01883	-0.01883	-0.01883	-0.021591	-0.01533	-0.015361	-0.021758	-0.021758	-0.089113	-0.068373	-0.308770	-0.062990	GRP_0	0
2	-0.025421	-0.026586	-0.018833	-0.015403	-0.015244	-0.017794	-0.015398	-0.028546	-0.026633	-0.134593	-0.133797	-0.035485	-0.014615	-0.021332	-0.028877	-0.018748	-0.015375	-0.015375	-0.199857	-0.052115	-0.031898	-0.015403	-0.014269	-0.015355	-0.038902	-0.024306	-0.020258	-0.020395	-0.029478	-0.014974	-0.01388	-0.015373	-0.06417	-0.015273	-0.0399	-0.060723	-0.015301	-0.015352	-0.040026	-0.014754	...	-0.015244	-0.015388	-0.021758	-0.021758	-0.015403	-0.015403	-0.01883	-0.015403	-0.01883	-0.01883	-0.01883	-0.01883	-0.015293	-0.01883	-0.01883	-0.033862	-0.015339	-0.015403	-0.015403	-0.015403	-0.021758	-0.026338	-0.015403	-0.025995	-0.018812	-0.01883	-0.01883	-0.01883	-0.01883	-0.021591	-0.01533	-0.015361	-0.021758	-0.021758	-0.535572	-0.509431	-1.740291	-0.515163	GRP_0	0
3	-0.025421	-0.026586	-0.018833	-0.015403	-0.015244	-0.017794	-0.015398	-0.028546	-0.026633	-0.134593	-0.133797	-0.035485	-0.014615	-0.021332	-0.028877	-0.018748	-0.015375	-0.015375	-0.199857	-0.052115	-0.031898	-0.015403	-0.014269	-0.015355	-0.038902	-0.024306	-0.020258	-0.020395	-0.029478	-0.014974	-0.01388	-0.015373	-0.06417	-0.015273	-0.0399	-0.060723	-0.015301	-0.015352	-0.040026	-0.014754	...	-0.015244	-0.015388	-0.021758	-0.021758	-0.015403	-0.015403	-0.01883	-0.015403	-0.01883	-0.01883	-0.01883	-0.01883	-0.015293	-0.01883	-0.01883	-0.033862	-0.015339	-0.015403	-0.015403	-0.015403	-0.021758	-0.026338	-0.015403	-0.025995	-0.018812	-0.01883	-0.01883	-0.01883	-0.01883	-0.021591	-0.01533	-0.015361	-0.021758	-0.021758	-0.447744	-0.399166	-1.440353	-0.402120	GRP_0	0
4	-0.025421	-0.026586	-0.018833	-0.015403	-0.015244	-0.017794	-0.015398	-0.028546	-0.026633	-0.134593	-0.133797	-0.035485	-0.014615	-0.021332	-0.028877	-0.018748	-0.015375	-0.015375	-0.199857	-0.052115	-0.031898	-0.015403	-0.014269	-0.015355	-0.038902	-0.024306	-0.020258	-0.020395	-0.029478	-0.014974	-0.01388	-0.015373	-0.06417	-0.015273	-0.0399	-0.060723	-0.015301	-0.015352	-0.040026	-0.014754	...	-0.015244	-0.015388	-0.021758	-0.021758	-0.015403	-0.015403	-0.01883	-0.015403	-0.01883	-0.01883	-0.01883	-0.01883	-0.015293	-0.01883	-0.01883	-0.033862	-0.015339	-0.015403	-0.015403	-0.015403	-0.021758	-0.026338	-0.015403	-0.025995	-0.018812	-0.01883	-0.01883	-0.01883	-0.01883	-0.021591	-0.01533	-0.015361	-0.021758	-0.021758	-0.557529	-0.564563	-0.990446	-0.571684	GRP_0	0

5 rows × 12276 columns

#Label encoding non-vectorized dataframe
le = preprocessing.LabelEncoder()
ticket_df['Assignment group'] = ticket_df['Assignment group'].astype(str)
ticket_df["LabelEncodings"] = le.fit_transform(ticket_df['Assignment group'])
y_classes_len = len(le.classes_)
le.classes_
print("No. of unique target group", y_classes_len)
ticket_df.head()

No. of unique target group 68

	Short description	Description	Caller	Assignment group	language	cleaned_description	descr_len	num_wds	avg_word	uniq_wds	tokenized_desc	LabelEncodings
0	login issue	-verified user details.(employee# & manager na...	spxjnwir pjlcoqds	GRP_0	en	login issue verified user detail employee mana...	125	18	6.000000	18	login issue verified user detail employee mana...	0
1	outlook	\r\n\r\nreceived from: hmjdrvpb.komuaywn@gmail...	hmjdrvpb komuaywn	GRP_0	en	outlook team meeting skype etc appearing calen...	75	11	5.909091	11	outlook team meeting skype etc appearing calen...	0
2	cant log in to vpn	\r\n\r\nreceived from: eylqgodm.ybqkwiam@gmail...	eylqgodm ybqkwiam	GRP_0	en	cannot log vpn	14	3	4.000000	3	cannot log vpn	0
3	unable to access hr_tool page	unable to access hr_tool page	xbkucsvz gcpydteq	GRP_0	en	unable access hr tool page	26	5	4.400000	5	unable access hr tool page	0
4	skype error	skype error	owlgqjme qhcozdfx	GRP_0	no	skype error	11	2	5.000000	2	skype error	0

Modelling

List of Machine Learning models to try out:

Statistical ML Models

• Multinomial Logistic Regression
• Support Vector Machine
• Stochastic Gradient Descent
• Multinomial Naive Bayes
• K Nearest neighbor
• Random Forest
• Decision trees
• Xgboost

Neural Network Models

• Deep Neural Network
• Convolutional Neural Network
• Recurrent Neural Network
• Recurrent Convolutional Neural Network
• LSTM

Creating Model with complete imbalance dataset

Statistical ML Models

bottom_20 = ticket_df['LabelEncodings'].value_counts().nsmallest(20).reset_index()
bottom_20

	index	LabelEncodings
0	56	1
1	64	2
2	49	2
3	65	2
4	52	2
5	62	2
6	51	3
7	31	3
8	58	3
9	61	3
10	53	3
11	26	4
12	60	4
13	37	5
14	40	6
15	54	6
16	43	6
17	50	8
18	46	8
19	47	9

bottom_20 = df_tfidf_z['LabelEncodings'].value_counts().nsmallest(20).reset_index()
bottom_20

	index	LabelEncodings
0	56	1
1	64	2
2	49	2
3	65	2
4	52	2
5	62	2
6	51	3
7	31	3
8	58	3
9	61	3
10	53	3
11	26	4
12	60	4
13	37	5
14	40	6
15	54	6
16	43	6
17	50	8
18	46	8
19	47	9

# Remove label encoded target group with one ticket each
ticket_df.drop(ticket_df[(ticket_df['LabelEncodings'] == 56)].index , inplace=True)
df_tfidf_z.drop(df_tfidf_z[(df_tfidf_z['LabelEncodings'] == 56)].index , inplace=True)

#Store the accuracy results for each model in a dataframe for final comparison
df_resLog = pd.DataFrame()
n_classes = len(ticket_df["LabelEncodings"].unique())
print(n_classes)

67

from sklearn.metrics import precision_recall_curve
# A class that logs the time
class Timer():
    '''
    A generic class to log the time
    '''
    def __init__(self):
        self.start_ts = None
    def start(self):
        self.start_ts = time()
    def stop(self):
        return 'Time taken: %2fs' % (time()-self.start_ts)
    
timer = Timer()

# A method to train and test the model
def run_classification(model_name, estimator, X_train, X_test, y_train, y_test, arch_name=None, pipelineRequired=True, isDeepModel=False):
    timer.start()
    # train the model
    clf = estimator

    if pipelineRequired :
        clf = Pipeline([('vect', CountVectorizer()),
                     ('tfidf', TfidfTransformer()),
                     ('clf', estimator),
                     ])
      
    if isDeepModel :
        clf.fit(X_train, y_train, validation_data=(X_test, y_test),epochs=10, batch_size=128,verbose=1,callbacks=call_backs(arch_name))
        # predict from the classifier
        y_pred = clf.predict(X_test)
        y_pred = np.argmax(y_pred, axis=1)
        y_train_pred = clf.predict(X_train)
        y_train_pred = np.argmax(y_train_pred, axis=1)
    else :
        print("start")
        clf.fit(X_train, y_train)
        # predict from the claffier
        y_pred = clf.predict(X_test)
        y_train_pred = clf.predict(X_train)
        y_pred_prob = clf.predict_proba(X_test)
    
    print('Estimator:', clf)
    print('='*80)
    train_acc = accuracy_score(y_train,y_train_pred)
    print('Training accuracy: ', (train_acc))
    test_acc = accuracy_score(y_test, y_pred)
    print('Testing accuracy: ', (test_acc))
    print('='*80)
    print('Confusion matrix:\n ', (confusion_matrix(y_test, y_pred)))
    print('='*80)
    print('Classification report:\n ', (classification_report(y_test, y_pred)))
    f1 = f1_score(y_pred, y_test, average='weighted', labels=np.unique(y_pred))
    print("F1 score: ",f1)
    print(timer.stop(), 'to run the model')

    # ROC curve
    fpr = dict()
    tpr = dict()
    thresh = dict()
    for i in range(n_classes):
      fpr[i], tpr[i], thresh[i] = roc_curve(y_test.ravel(), y_pred_prob[:,i], pos_label=i)

    # plotting   
    plt.figure(figsize=(20,6))
    plt.subplot(1, 2, 1) 
    plt.plot(fpr[0], tpr[0], linestyle='--',color='orange', label='Class 0')
    plt.plot(fpr[66], tpr[66], linestyle='--',color='green', label='Class 66')
    plt.plot(fpr[17], tpr[17], linestyle='--',color='blue', label='Class 17')
    plt.plot(fpr[4], tpr[4], linestyle='--',color='yellow', label='Class 4')
    plt.plot([0, 1], [0, 1], linestyle='--', color='red', label='ideal')
    plt.title('Multiclass ROC curve')
    plt.xlabel('False Positive Rate')
    plt.ylabel('True Positive rate')
    plt.legend(loc='best')

    # Precision Recall curve
    precision = dict()
    recall = dict()
    thres = dict()
    for i in range(n_classes):
      precision[i], recall[i], thres[i] = precision_recall_curve(y_test.ravel(), y_pred_prob[:,i], pos_label=i)

    # plotting   
    plt.subplot(1, 2, 2) 
    plt.plot(recall[0], precision[0], linestyle='--',color='orange', label='Class 0')
    plt.plot(recall[66], precision[66], linestyle='--',color='green', label='Class 66')
    plt.plot(recall[17], precision[17], linestyle='--',color='blue', label='Class 17')
    plt.plot(recall[4], precision[4],color='yellow', label='Class 4')
    plt.title('Multiclass Precision Recall curve')
    plt.xlabel('recall')
    plt.ylabel('precision')
    plt.legend(loc='best')

    t = re.findall('\d*\.?\d+',(timer.stop()))
    df_results = pd.DataFrame()
    df_tempResults = pd.DataFrame({'Model':[model_name], 'Train Accuracy': [train_acc], 'Test Accuracy': [test_acc], 'F1 Score': [f1],  'Time Taken (sec)': [t[0]]})
    df_results = pd.concat([df_results, df_tempResults])
    return df_results

# Unvectorized cleansed data
X = ticket_df.tokenized_desc
Y = ticket_df["LabelEncodings"]
X_train, X_test, y_train, y_test = train_test_split(X, 
                                                    Y	, 
                                                    test_size=0.20, 
                                                    random_state=42, stratify = ticket_df["LabelEncodings"])
print('Shape of the training set:', X_train.shape, X_test.shape)
print('Shape of the test set:', y_train.shape, y_test.shape)

Shape of the training set: (6744,) (1687,)
Shape of the test set: (6744,) (1687,)

# Vectorized dataframe without word count features
X_v = df_tfidf_z.iloc[:,:-6]
Y_v = df_tfidf_z['LabelEncodings']
input = 'One-hot encoded target with no. words, avg number, len, unique word no'
X_train_v, X_test_v, y_train_v, y_test_v = train_test_split(X_v, 
                                                    Y_v, 
                                                    test_size=0.20, 
                                                    random_state=42)
print('Shape of the training set:', X_train_v.shape, X_test_v.shape)
print('Shape of the test set:', y_train_v.shape, y_test_v.shape)

Shape of the training set: (6744, 12270) (1687, 12270)
Shape of the test set: (6744,) (1687,)

# Vectorized dataframe with word count features
X_vf = df_tfidf_z.iloc[:,:-2]
Y_vf = df_tfidf_z['LabelEncodings']
input = 'One-hot encoded target with no. words, avg number, len, unique word no'
X_train_vf, X_test_vf, y_train_vf, y_test_vf = train_test_split(X_vf, 
                                                    Y_vf, 
                                                    test_size=0.20, 
                                                    random_state=42)
print('Shape of the training set:', X_train_vf.shape, X_test_vf.shape)
print('Shape of the test set:', y_train_vf.shape, y_test_vf.shape)

Shape of the training set: (6744, 12274) (1687, 12274)
Shape of the test set: (6744,) (1687,)

# No of unique targets present in the train and test set 
Ytrain = pd.DataFrame()
Ytrain['Assignment group'] = le.inverse_transform(y_train)
Ytest = pd.DataFrame()
Ytest['Assignment group'] = le.inverse_transform(y_test)

train_test_df = pd.DataFrame()
train_test_df['Unique Targets'] = [len(Ytrain['Assignment group'].unique()), len(Ytest['Assignment group'].unique())]
train_test_df.index = ['Train', 'Test']
train_test_df.transpose()

	Train	Test
Unique Targets	67	62

The no. of unique targets in train is not the same with test due to high imbalance in the data, having as low as 2 tickets in few groups before splitting.

# Train set target distribution
configure_plotly_browser_state()
Ytrain['Assignment group'].iplot(kind='hist', xTitle='Assignment Group', yTitle='count', colorscale='-orrd', title='Records by rare Assignment Groups- Histogram')

plt.show()

# Test set target distribution
configure_plotly_browser_state()
Ytest['Assignment group'].iplot(kind='hist', xTitle='Assignment Group', yTitle='count', colorscale='-orrd', title='Records by rare Assignment Groups- Histogram')

plt.show()

Multinomial Logistic Regression

df_LR_Results = pd.DataFrame(columns=['Model','Train Accuracy','Test Accuracy','F1 Score','Time Taken (sec)'])

model_name = 'Multinomial Logistic Regression'
df_temp = run_classification(model_name, LogisticRegression(multi_class='multinomial', solver='lbfgs'), X_train, X_test, y_train, y_test)
df_LR_Results = pd.concat([df_LR_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)]=df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=Non...
                ('tfidf',
                 TfidfTransformer(norm='l2', smooth_idf=True,
                                  sublinear_tf=False, use_idf=True)),
                ('clf',
                 LogisticRegression(C=1.0, class_weight=None, dual=False,
                                    fit_intercept=True, intercept_scaling=1,
                                    l1_ratio=None, max_iter=100,
                                    multi_class='multinomial', n_jobs=None,
                                    penalty='l2', random_state=None,
                                    solver='lbfgs', tol=0.0001, verbose=0,
                                    warm_start=False))],
         verbose=False)
================================================================================
Training accuracy:  0.663552787663108
Testing accuracy:  0.6218138707765264
================================================================================
Confusion matrix:
  [[783   0   0 ...   0   0   0]
 [  1   0   0 ...   0   3   0]
 [ 15   0   1 ...   0  12   0]
 ...
 [ 10   0   0 ...   4   0   0]
 [  4   0   0 ...   0 125   0]
 [ 10   0   0 ...   0  34   3]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.63      0.99      0.77       792
           1       0.00      0.00      0.00         6
           2       0.50      0.04      0.07        28
           3       0.00      0.00      0.00         6
           4       0.57      0.31      0.41        51
           5       0.59      0.45      0.51        29
           6       0.62      0.33      0.43        24
           7       0.00      0.00      0.00         8
           8       0.00      0.00      0.00        17
           9       0.93      0.88      0.90        16
          10       0.60      0.17      0.26        18
          11       0.67      0.09      0.16        43
          12       0.66      0.40      0.49        48
          13       0.00      0.00      0.00         7
          14       0.00      0.00      0.00         6
          15       0.00      0.00      0.00         6
          16       0.00      0.00      0.00         5
          17       0.86      0.68      0.76        56
          18       0.50      0.17      0.26        23
          19       0.00      0.00      0.00        11
          20       0.00      0.00      0.00         4
          21       0.00      0.00      0.00         9
          22       0.67      0.11      0.18        19
          23       0.38      0.07      0.12        40
          24       0.00      0.00      0.00         7
          25       0.00      0.00      0.00        14
          26       0.00      0.00      0.00         1
          27       0.71      0.24      0.36        21
          28       0.00      0.00      0.00        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       0.00      0.00      0.00         4
          33       0.00      0.00      0.00        20
          34       0.00      0.00      0.00         9
          35       0.00      0.00      0.00         8
          36       0.00      0.00      0.00         7
          37       0.00      0.00      0.00         1
          38       0.00      0.00      0.00         3
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       0.00      0.00      0.00         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.00      0.00      0.00        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         1
          53       0.00      0.00      0.00         1
          54       0.00      0.00      0.00         1
          55       0.44      0.11      0.18        36
          57       0.00      0.00      0.00         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       0.00      0.00      0.00         1
          63       1.00      0.29      0.44        14
          66       0.54      0.95      0.69       132
          67       1.00      0.06      0.11        50

    accuracy                           0.62      1687
   macro avg       0.19      0.10      0.11      1687
weighted avg       0.54      0.62      0.52      1687

F1 score:  0.7228415587532678
Time taken: 18.707438s to run the model

model_name = 'Multinomial Logistic Regression with b-weights'
df_temp = run_classification(model_name, LogisticRegression(multi_class='multinomial', solver='lbfgs',class_weight='balanced'), X_train, X_test, y_train, y_test)
df_LR_Results = pd.concat([df_LR_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)]=df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=Non...
                ('tfidf',
                 TfidfTransformer(norm='l2', smooth_idf=True,
                                  sublinear_tf=False, use_idf=True)),
                ('clf',
                 LogisticRegression(C=1.0, class_weight='balanced', dual=False,
                                    fit_intercept=True, intercept_scaling=1,
                                    l1_ratio=None, max_iter=100,
                                    multi_class='multinomial', n_jobs=None,
                                    penalty='l2', random_state=None,
                                    solver='lbfgs', tol=0.0001, verbose=0,
                                    warm_start=False))],
         verbose=False)
================================================================================
Training accuracy:  0.5029655990510083
Testing accuracy:  0.3787788974510966
================================================================================
Confusion matrix:
  [[244   0   0 ...  38   0   0]
 [  0   1   0 ...   0   0   0]
 [  0   0  10 ...   0   0   0]
 ...
 [  0   0   0 ...   0   0   0]
 [  0   0   0 ...   0  35   0]
 [  0   0   0 ...   0   0   3]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.96      0.31      0.47       792
           1       0.50      0.17      0.25         6
           2       1.00      0.36      0.53        28
           3       0.12      0.33      0.17         6
           4       0.52      0.33      0.40        51
           5       0.49      0.62      0.55        29
           6       0.29      0.50      0.36        24
           7       0.30      0.38      0.33         8
           8       0.29      0.71      0.41        17
           9       0.41      1.00      0.58        16
          10       0.31      0.61      0.42        18
          11       0.22      0.35      0.27        43
          12       0.34      0.52      0.41        48
          13       0.00      0.00      0.00         7
          14       0.11      0.33      0.17         6
          15       0.28      0.83      0.42         6
          16       0.29      1.00      0.45         5
          17       0.78      0.71      0.75        56
          18       0.29      0.70      0.41        23
          19       0.09      0.64      0.16        11
          20       0.00      0.00      0.00         4
          21       0.19      0.33      0.24         9
          22       0.44      0.58      0.50        19
          23       0.18      0.28      0.22        40
          24       0.75      0.43      0.55         7
          25       0.15      0.50      0.23        14
          26       0.00      0.00      0.00         1
          27       0.41      0.57      0.48        21
          28       0.12      0.42      0.19        12
          29       0.29      0.67      0.40         3
          30       0.00      0.00      0.00         3
          31       1.00      1.00      1.00         1
          32       1.00      0.50      0.67         4
          33       0.31      0.60      0.41        20
          34       0.20      0.67      0.31         9
          35       0.42      1.00      0.59         8
          36       0.16      0.43      0.23         7
          37       0.25      1.00      0.40         1
          38       0.20      0.33      0.25         3
          39       0.23      0.43      0.30         7
          40       0.33      1.00      0.50         1
          41       0.12      0.80      0.21         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.12      0.69      0.21        26
          45       0.12      0.67      0.21         3
          46       1.00      0.50      0.67         2
          47       0.00      0.00      0.00         2
          48       0.20      0.50      0.29         2
          50       0.40      1.00      0.57         2
          51       0.00      0.00      0.00         1
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.40      0.06      0.10        36
          57       0.40      0.80      0.53         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       1.00      1.00      1.00         1
          61       0.25      1.00      0.40         1
          62       0.00      0.00      0.00         0
          63       0.31      0.79      0.44        14
          65       0.00      0.00      0.00         0
          66       0.92      0.27      0.41       132
          67       1.00      0.06      0.11        50

    accuracy                           0.38      1687
   macro avg       0.34      0.47      0.33      1687
weighted avg       0.71      0.38      0.41      1687

F1 score:  0.3429446404041265
Time taken: 11.590784s to run the model

model_name = 'Multinomial Logistic Regression without word count features'
df_temp = run_classification(model_name, LogisticRegression(multi_class='multinomial', solver='lbfgs'), X_train_v, X_test_v, y_train_v, y_test_v, arch_name= None, pipelineRequired=False)
df_LR_Results = pd.concat([df_LR_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)]=df_temp

start
Estimator: LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,
                   intercept_scaling=1, l1_ratio=None, max_iter=100,
                   multi_class='multinomial', n_jobs=None, penalty='l2',
                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,
                   warm_start=False)
================================================================================
Training accuracy:  0.9297153024911032
Testing accuracy:  0.6419679905157084
================================================================================
Confusion matrix:
  [[660   0   3 ...   6   1   1]
 [  0   0   0 ...   0   2   0]
 [  2   0  10 ...   0  12   0]
 ...
 [  4   0   0 ...   8   0   0]
 [  1   0   0 ...   0 134   2]
 [  0   0   1 ...   0  27   8]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.82      0.82      0.82       803
           1       0.00      0.00      0.00         3
           2       0.48      0.32      0.38        31
           3       0.00      0.00      0.00         9
           4       0.48      0.51      0.49        39
           5       0.52      0.46      0.49        28
           6       0.33      0.40      0.36        20
           7       0.50      0.33      0.40         9
           8       0.33      0.40      0.36        15
           9       0.88      0.88      0.88        17
          10       0.27      0.32      0.29        19
          11       0.48      0.38      0.42        56
          12       0.58      0.60      0.59        48
          13       0.00      0.00      0.00         7
          14       0.33      0.14      0.20         7
          15       0.25      0.50      0.33         4
          16       0.40      0.50      0.44         4
          17       0.70      0.81      0.75        47
          18       0.54      0.44      0.48        32
          19       0.38      0.25      0.30        12
          20       0.00      0.00      0.00         1
          21       0.18      0.18      0.18        11
          22       0.62      0.50      0.55        16
          23       0.34      0.37      0.35        38
          24       0.25      0.12      0.17         8
          25       0.40      0.40      0.40        15
          26       1.00      1.00      1.00         1
          27       0.38      0.36      0.37        22
          28       0.20      0.15      0.17        13
          29       0.50      0.50      0.50         2
          30       0.33      0.17      0.22         6
          32       1.00      0.50      0.67         2
          33       0.25      0.27      0.26        15
          34       0.50      0.18      0.27        11
          35       0.60      0.43      0.50         7
          36       0.00      0.00      0.00         3
          37       0.00      0.00      0.00         1
          38       0.67      0.40      0.50         5
          39       0.40      0.29      0.33         7
          40       0.00      0.00      0.00         1
          41       0.67      0.29      0.40         7
          42       0.36      1.00      0.53         8
          43       0.00      0.00      0.00         3
          44       0.00      0.00      0.00        22
          45       0.00      0.00      0.00         2
          46       1.00      1.00      1.00         1
          47       0.00      0.00      0.00         1
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         2
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.44      0.24      0.31        29
          57       1.00      0.57      0.73         7
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       1.00      1.00      1.00         1
          63       0.50      0.62      0.55        13
          66       0.57      0.91      0.70       148
          67       0.57      0.21      0.31        38

    accuracy                           0.64      1687
   macro avg       0.38      0.34      0.34      1687
weighted avg       0.64      0.64      0.63      1687

F1 score:  0.6545086593423995
Time taken: 114.398176s to run the model

model_name = 'Multinomial Logistic Regression with word count features'
df_temp = run_classification(model_name, LogisticRegression(multi_class='multinomial', solver='lbfgs'), X_train_vf, X_test_vf, y_train_vf, y_test_vf, arch_name= None, pipelineRequired=False)
df_LR_Results = pd.concat([df_LR_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)]=df_temp

start
Estimator: LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,
                   intercept_scaling=1, l1_ratio=None, max_iter=100,
                   multi_class='multinomial', n_jobs=None, penalty='l2',
                   random_state=None, solver='lbfgs', tol=0.0001, verbose=0,
                   warm_start=False)
================================================================================
Training accuracy:  0.9292704626334519
Testing accuracy:  0.6461173681090694
================================================================================
Confusion matrix:
  [[667   0   3 ...   6   1   1]
 [  0   0   0 ...   0   2   0]
 [  2   0  12 ...   0  12   0]
 ...
 [  4   0   0 ...   8   0   0]
 [  1   0   0 ...   0 135   2]
 [  0   0   1 ...   0  27   8]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.82      0.83      0.83       803
           1       0.00      0.00      0.00         3
           2       0.52      0.39      0.44        31
           3       0.00      0.00      0.00         9
           4       0.50      0.51      0.51        39
           5       0.50      0.46      0.48        28
           6       0.35      0.40      0.37        20
           7       0.50      0.33      0.40         9
           8       0.33      0.40      0.36        15
           9       0.88      0.88      0.88        17
          10       0.29      0.32      0.30        19
          11       0.45      0.34      0.39        56
          12       0.57      0.62      0.59        48
          13       0.00      0.00      0.00         7
          14       0.33      0.14      0.20         7
          15       0.25      0.50      0.33         4
          16       0.40      0.50      0.44         4
          17       0.72      0.81      0.76        47
          18       0.54      0.44      0.48        32
          19       0.43      0.25      0.32        12
          20       0.00      0.00      0.00         1
          21       0.18      0.18      0.18        11
          22       0.67      0.50      0.57        16
          23       0.37      0.37      0.37        38
          24       0.25      0.12      0.17         8
          25       0.46      0.40      0.43        15
          26       1.00      1.00      1.00         1
          27       0.36      0.36      0.36        22
          28       0.18      0.15      0.17        13
          29       0.50      0.50      0.50         2
          30       0.33      0.17      0.22         6
          32       1.00      0.50      0.67         2
          33       0.31      0.27      0.29        15
          34       0.50      0.18      0.27        11
          35       0.60      0.43      0.50         7
          36       0.00      0.00      0.00         3
          37       0.00      0.00      0.00         1
          38       0.67      0.40      0.50         5
          39       0.40      0.29      0.33         7
          40       0.00      0.00      0.00         1
          41       0.67      0.29      0.40         7
          42       0.29      0.62      0.40         8
          43       0.00      0.00      0.00         3
          44       0.00      0.00      0.00        22
          45       0.00      0.00      0.00         2
          46       1.00      1.00      1.00         1
          47       0.00      0.00      0.00         1
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         2
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.47      0.28      0.35        29
          57       1.00      0.57      0.73         7
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       1.00      1.00      1.00         1
          63       0.47      0.62      0.53        13
          66       0.57      0.91      0.70       148
          67       0.57      0.21      0.31        38

    accuracy                           0.65      1687
   macro avg       0.38      0.34      0.34      1687
weighted avg       0.64      0.65      0.63      1687

F1 score:  0.6601124350194423
Time taken: 119.008312s to run the model

df_LR_Results.sort_values(by='Test Accuracy', ascending = False)

	Model	Train Accuracy	Test Accuracy	F1 Score	Time Taken (sec)
0	Multinomial Logistic Regression with word coun...	0.929270	0.646117	0.660112	119.157214
0	Multinomial Logistic Regression without word c...	0.929715	0.641968	0.654509	114.525980
0	Multinomial Logistic Regression	0.663553	0.621814	0.722842	18.808882
0	Multinomial Logistic Regression with b-weights	0.502966	0.378779	0.342945	11.690096

df_LR_Results.plot(kind='barh',x='Model', figsize=[15,10])

<matplotlib.axes._subplots.AxesSubplot at 0x7f70981c7050>

Support Vector Machine

model_name = 'SVM-RBF Kernal'
df_temp = run_classification(model_name, SVC(kernel='rbf', probability = True), X_train, X_test, y_train, y_test)
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)]=df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=None)),
                ('tfidf',
                 TfidfTransformer(norm='l2', smooth_idf=True,
                                  sublinear_tf=False, use_idf=True)),
                ('clf',
                 SVC(C=1.0, break_ties=False, cache_size=200, class_weight=None,
                     coef0=0.0, decision_function_shape='ovr', degree=3,
                     gamma='scale', kernel='rbf', max_iter=-1, probability=True,
                     random_state=None, shrinking=True, tol=0.001,
                     verbose=False))],
         verbose=False)
================================================================================
Training accuracy:  0.7937425860023725
Testing accuracy:  0.6200355660936574
================================================================================
Confusion matrix:
  [[788   0   0 ...   0   0   1]
 [  2   1   0 ...   0   3   0]
 [ 12   0   4 ...   0  12   0]
 ...
 [  9   0   0 ...   5   0   0]
 [  1   0   0 ...   0 126   0]
 [ 13   0   0 ...   0  34   3]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.61      0.99      0.76       792
           1       1.00      0.17      0.29         6
           2       1.00      0.14      0.25        28
           3       0.00      0.00      0.00         6
           4       0.70      0.27      0.39        51
           5       0.29      0.07      0.11        29
           6       0.67      0.33      0.44        24
           7       0.00      0.00      0.00         8
           8       0.00      0.00      0.00        17
           9       0.94      0.94      0.94        16
          10       0.75      0.17      0.27        18
          11       0.50      0.02      0.04        43
          12       0.71      0.35      0.47        48
          13       0.00      0.00      0.00         7
          14       0.00      0.00      0.00         6
          15       0.00      0.00      0.00         6
          16       0.00      0.00      0.00         5
          17       1.00      0.64      0.78        56
          18       0.67      0.26      0.38        23
          19       0.00      0.00      0.00        11
          20       0.00      0.00      0.00         4
          21       0.00      0.00      0.00         9
          22       1.00      0.11      0.19        19
          23       0.50      0.05      0.09        40
          24       0.00      0.00      0.00         7
          25       0.00      0.00      0.00        14
          26       0.00      0.00      0.00         1
          27       0.71      0.24      0.36        21
          28       0.00      0.00      0.00        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       0.00      0.00      0.00         4
          33       1.00      0.05      0.10        20
          34       0.00      0.00      0.00         9
          35       0.00      0.00      0.00         8
          36       0.00      0.00      0.00         7
          37       0.00      0.00      0.00         1
          38       0.00      0.00      0.00         3
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       0.00      0.00      0.00         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.00      0.00      0.00        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         1
          53       0.00      0.00      0.00         1
          54       0.00      0.00      0.00         1
          55       0.58      0.19      0.29        36
          57       0.00      0.00      0.00         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       0.00      0.00      0.00         1
          63       0.83      0.36      0.50        14
          66       0.56      0.95      0.70       132
          67       0.75      0.06      0.11        50

    accuracy                           0.62      1687
   macro avg       0.24      0.10      0.12      1687
weighted avg       0.56      0.62      0.52      1687

F1 score:  0.7232775002682313
Time taken: 126.292463s to run the model

model_name = 'SVM-RBF Kernal with b-weights'
df_temp = run_classification(model_name, SVC(kernel='rbf', probability = True, class_weight='balanced'), X_train, X_test, y_train, y_test)
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)]=df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=None)),
                ('tfidf',
                 TfidfTransformer(norm='l2', smooth_idf=True,
                                  sublinear_tf=False, use_idf=True)),
                ('clf',
                 SVC(C=1.0, break_ties=False, cache_size=200,
                     class_weight='balanced', coef0=0.0,
                     decision_function_shape='ovr', degree=3, gamma='scale',
                     kernel='rbf', max_iter=-1, probability=True,
                     random_state=None, shrinking=True, tol=0.001,
                     verbose=False))],
         verbose=False)
================================================================================
Training accuracy:  0.6316725978647687
Testing accuracy:  0.45168938944872555
================================================================================
Confusion matrix:
  [[396   0   0 ...   9   0   0]
 [  0   1   0 ...   0   0   0]
 [  1   0   5 ...   0   0   0]
 ...
 [  0   0   0 ...   0   0   0]
 [  0   0   0 ...   0  33   0]
 [  0   0   0 ...   0   0   3]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.93      0.50      0.65       792
           1       0.50      0.17      0.25         6
           2       0.83      0.18      0.29        28
           3       0.00      0.00      0.00         6
           4       0.55      0.43      0.48        51
           5       0.39      0.93      0.55        29
           6       0.31      0.50      0.38        24
           7       0.50      0.12      0.20         8
           8       0.38      0.65      0.48        17
           9       0.71      0.94      0.81        16
          10       0.46      0.33      0.39        18
          11       0.09      0.63      0.15        43
          12       0.39      0.60      0.48        48
          13       0.00      0.00      0.00         7
          14       1.00      0.17      0.29         6
          15       0.67      0.67      0.67         6
          16       0.43      0.60      0.50         5
          17       0.90      0.64      0.75        56
          18       0.35      0.78      0.48        23
          19       0.21      0.36      0.27        11
          20       0.00      0.00      0.00         4
          21       0.33      0.11      0.17         9
          22       0.65      0.58      0.61        19
          23       0.11      0.40      0.17        40
          24       1.00      0.43      0.60         7
          25       0.50      0.29      0.36        14
          26       0.00      0.00      0.00         1
          27       0.39      0.76      0.52        21
          28       0.36      0.33      0.35        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       0.50      0.50      0.50         4
          33       0.33      0.30      0.32        20
          34       0.13      0.22      0.17         9
          35       0.60      0.75      0.67         8
          36       0.00      0.00      0.00         7
          37       0.00      0.00      0.00         1
          38       0.00      0.00      0.00         3
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       0.08      0.40      0.13         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.12      0.65      0.20        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.00      0.00      0.00         2
          50       1.00      0.50      0.67         2
          51       0.00      0.00      0.00         1
          53       0.00      0.00      0.00         1
          54       0.00      0.00      0.00         1
          55       0.86      0.17      0.28        36
          57       0.00      0.00      0.00         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       1.00      1.00      1.00         1
          61       0.00      0.00      0.00         1
          63       0.67      0.71      0.69        14
          65       0.00      0.00      0.00         0
          66       0.89      0.25      0.39       132
          67       1.00      0.06      0.11        50

    accuracy                           0.45      1687
   macro avg       0.32      0.28      0.25      1687
weighted avg       0.72      0.45      0.50      1687

F1 score:  0.40465245950009504
Time taken: 209.476274s to run the model

Stochastic Gradient Descent

model_name = 'Stochastic Gradient Descent'
df_temp = run_classification(model_name, SGDClassifier(loss='modified_huber', penalty='l2',alpha=1e-3, random_state=42, max_iter=100, tol=None), X_train, X_test, y_train, y_test)
# df_SGD_Results = pd.concat([df_SGD_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)]=df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=Non...
                 SGDClassifier(alpha=0.001, average=False, class_weight=None,
                               early_stopping=False, epsilon=0.1, eta0=0.0,
                               fit_intercept=True, l1_ratio=0.15,
                               learning_rate='optimal', loss='modified_huber',
                               max_iter=100, n_iter_no_change=5, n_jobs=None,
                               penalty='l2', power_t=0.5, random_state=42,
                               shuffle=True, tol=None, validation_fraction=0.1,
                               verbose=0, warm_start=False))],
         verbose=False)
================================================================================
Training accuracy:  0.7620106761565836
Testing accuracy:  0.6544161232957914
================================================================================
Confusion matrix:
  [[773   0   1 ...   2   0   0]
 [  1   0   0 ...   0   3   0]
 [ 10   0   3 ...   0  12   0]
 ...
 [  7   0   0 ...   5   0   0]
 [  3   0   0 ...   0 126   0]
 [  8   0   0 ...   0  34   6]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.68      0.98      0.80       792
           1       0.00      0.00      0.00         6
           2       0.43      0.11      0.17        28
           3       0.00      0.00      0.00         6
           4       0.67      0.39      0.49        51
           5       0.57      0.69      0.62        29
           6       0.67      0.42      0.51        24
           7       0.00      0.00      0.00         8
           8       0.00      0.00      0.00        17
           9       0.83      0.94      0.88        16
          10       0.64      0.39      0.48        18
          11       0.60      0.14      0.23        43
          12       0.56      0.40      0.46        48
          13       0.00      0.00      0.00         7
          14       1.00      0.17      0.29         6
          15       0.00      0.00      0.00         6
          16       0.00      0.00      0.00         5
          17       0.80      0.73      0.77        56
          18       0.75      0.39      0.51        23
          19       0.00      0.00      0.00        11
          20       0.00      0.00      0.00         4
          21       0.00      0.00      0.00         9
          22       0.58      0.37      0.45        19
          23       0.50      0.20      0.29        40
          24       1.00      0.14      0.25         7
          25       0.00      0.00      0.00        14
          26       0.00      0.00      0.00         1
          27       0.57      0.38      0.46        21
          28       0.00      0.00      0.00        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       0.00      0.00      0.00         4
          33       0.50      0.10      0.17        20
          34       1.00      0.11      0.20         9
          35       1.00      0.50      0.67         8
          36       0.00      0.00      0.00         7
          37       0.00      0.00      0.00         1
          38       0.00      0.00      0.00         3
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       1.00      0.20      0.33         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.00      0.00      0.00        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         1
          53       0.00      0.00      0.00         1
          54       0.00      0.00      0.00         1
          55       0.50      0.31      0.38        36
          57       0.00      0.00      0.00         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       0.00      0.00      0.00         1
          63       0.62      0.36      0.45        14
          66       0.55      0.95      0.70       132
          67       0.67      0.12      0.20        50

    accuracy                           0.65      1687
   macro avg       0.27      0.15      0.17      1687
weighted avg       0.58      0.65      0.58      1687

F1 score:  0.7331472222209854
Time taken: 5.053100s to run the model

model_name = 'Stochastic Gradient Descent with b-weights'
df_temp = run_classification(model_name, SGDClassifier(loss='modified_huber', penalty='l2',alpha=1e-3, random_state=42, max_iter=100, tol=None,class_weight='balanced'), X_train, X_test, y_train, y_test)
# df_SGD_Results = pd.concat([df_SGD_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)]=df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=Non...
                 SGDClassifier(alpha=0.001, average=False,
                               class_weight='balanced', early_stopping=False,
                               epsilon=0.1, eta0=0.0, fit_intercept=True,
                               l1_ratio=0.15, learning_rate='optimal',
                               loss='modified_huber', max_iter=100,
                               n_iter_no_change=5, n_jobs=None, penalty='l2',
                               power_t=0.5, random_state=42, shuffle=True,
                               tol=None, validation_fraction=0.1, verbose=0,
                               warm_start=False))],
         verbose=False)
================================================================================
Training accuracy:  0.6803084223013048
Testing accuracy:  0.4919976289270895
================================================================================
Confusion matrix:
  [[438   2   0 ...   9   0   1]
 [  0   1   0 ...   0   0   0]
 [  0   0  11 ...   0   0   0]
 ...
 [  0   0   0 ...   0   0   0]
 [  0   0   0 ...   0  43   0]
 [  0   0   0 ...   0   0   7]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.95      0.55      0.70       792
           1       0.20      0.17      0.18         6
           2       0.79      0.39      0.52        28
           3       0.07      0.17      0.10         6
           4       0.53      0.33      0.41        51
           5       0.55      0.62      0.58        29
           6       0.41      0.54      0.46        24
           7       0.25      0.50      0.33         8
           8       0.29      0.71      0.41        17
           9       0.52      1.00      0.68        16
          10       0.32      0.56      0.41        18
          11       0.34      0.30      0.32        43
          12       0.55      0.46      0.50        48
          13       0.00      0.00      0.00         7
          14       0.10      0.33      0.15         6
          15       0.19      0.67      0.30         6
          16       0.26      1.00      0.42         5
          17       0.77      0.82      0.79        56
          18       0.40      0.74      0.52        23
          19       0.10      0.36      0.15        11
          20       0.00      0.00      0.00         4
          21       0.17      0.33      0.22         9
          22       0.46      0.63      0.53        19
          23       0.30      0.30      0.30        40
          24       0.39      1.00      0.56         7
          25       0.17      0.29      0.22        14
          26       0.00      0.00      0.00         1
          27       0.46      0.57      0.51        21
          28       0.22      0.42      0.29        12
          29       0.20      0.67      0.31         3
          30       0.00      0.00      0.00         3
          31       0.50      1.00      0.67         1
          32       0.40      0.50      0.44         4
          33       0.36      0.45      0.40        20
          34       0.23      0.67      0.34         9
          35       0.32      0.88      0.47         8
          36       0.17      0.43      0.24         7
          37       0.33      1.00      0.50         1
          38       0.25      0.33      0.29         3
          39       0.30      0.43      0.35         7
          40       0.08      1.00      0.15         1
          41       0.13      0.80      0.23         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.50      0.04      0.07        26
          45       0.12      0.33      0.18         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.15      1.00      0.27         2
          50       0.25      0.50      0.33         2
          51       0.00      0.00      0.00         1
          52       0.00      0.00      0.00         0
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.57      0.22      0.32        36
          57       0.33      0.80      0.47         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       1.00      1.00      1.00         1
          61       1.00      1.00      1.00         1
          62       0.00      0.00      0.00         0
          63       0.42      0.79      0.55        14
          65       0.00      0.00      0.00         0
          66       0.93      0.33      0.48       132
          67       0.70      0.14      0.23        50

    accuracy                           0.49      1687
   macro avg       0.31      0.45      0.32      1687
weighted avg       0.72      0.49      0.55      1687

F1 score:  0.4378647024280598
Time taken: 4.927700s to run the model

Multinomial Naive Bayes

model_name = 'Multinomial Naive Bayes'
df_temp = run_classification(model_name, MultinomialNB(), X_train, X_test, y_train, y_test)
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)]  =df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=None)),
                ('tfidf',
                 TfidfTransformer(norm='l2', smooth_idf=True,
                                  sublinear_tf=False, use_idf=True)),
                ('clf',
                 MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True))],
         verbose=False)
================================================================================
Training accuracy:  0.5682087781731909
Testing accuracy:  0.5548310610551275
================================================================================
Confusion matrix:
  [[792   0   0 ...   0   0   0]
 [  2   0   0 ...   0   3   0]
 [ 13   0   0 ...   0  15   0]
 ...
 [ 14   0   0 ...   0   0   0]
 [  7   0   0 ...   0 125   0]
 [ 13   0   0 ...   0  37   0]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.56      1.00      0.72       792
           1       0.00      0.00      0.00         6
           2       0.00      0.00      0.00        28
           3       0.00      0.00      0.00         6
           4       0.50      0.08      0.14        51
           5       0.00      0.00      0.00        29
           6       0.00      0.00      0.00        24
           7       0.00      0.00      0.00         8
           8       0.00      0.00      0.00        17
           9       0.00      0.00      0.00        16
          10       0.00      0.00      0.00        18
          11       0.00      0.00      0.00        43
          12       0.45      0.10      0.17        48
          13       0.00      0.00      0.00         7
          14       0.00      0.00      0.00         6
          15       0.00      0.00      0.00         6
          16       0.00      0.00      0.00         5
          17       1.00      0.18      0.30        56
          18       0.00      0.00      0.00        23
          19       0.00      0.00      0.00        11
          20       0.00      0.00      0.00         4
          21       0.00      0.00      0.00         9
          22       0.00      0.00      0.00        19
          23       0.00      0.00      0.00        40
          24       0.00      0.00      0.00         7
          25       0.00      0.00      0.00        14
          26       0.00      0.00      0.00         1
          27       0.00      0.00      0.00        21
          28       0.00      0.00      0.00        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       0.00      0.00      0.00         4
          33       0.00      0.00      0.00        20
          34       0.00      0.00      0.00         9
          35       0.00      0.00      0.00         8
          36       0.00      0.00      0.00         7
          37       0.00      0.00      0.00         1
          38       0.00      0.00      0.00         3
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       0.00      0.00      0.00         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.00      0.00      0.00        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         1
          53       0.00      0.00      0.00         1
          54       0.00      0.00      0.00         1
          55       0.00      0.00      0.00        36
          57       0.00      0.00      0.00         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       0.00      0.00      0.00         1
          63       0.00      0.00      0.00        14
          66       0.53      0.95      0.68       132
          67       0.00      0.00      0.00        50

    accuracy                           0.55      1687
   macro avg       0.05      0.04      0.03      1687
weighted avg       0.36      0.55      0.41      1687

F1 score:  0.7016264330788983
Time taken: 0.271396s to run the model

K Nearest neighbor

model_name = 'K Nearest neighbor'
df_temp = run_classification(model_name, KNeighborsClassifier(), X_train, X_test, y_train, y_test)
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)]=df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=None)),
                ('tfidf',
                 TfidfTransformer(norm='l2', smooth_idf=True,
                                  sublinear_tf=False, use_idf=True)),
                ('clf',
                 KNeighborsClassifier(algorithm='auto', leaf_size=30,
                                      metric='minkowski', metric_params=None,
                                      n_jobs=None, n_neighbors=5, p=2,
                                      weights='uniform'))],
         verbose=False)
================================================================================
Training accuracy:  0.6933570581257414
Testing accuracy:  0.6336692353289863
================================================================================
Confusion matrix:
  [[771   0   0 ...   1   0   0]
 [  1   1   0 ...   0   3   0]
 [  5   0   5 ...   0  15   0]
 ...
 [ 10   0   0 ...   4   0   0]
 [  3   0   0 ...   0 121   0]
 [  8   0   0 ...   0  34   5]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.68      0.97      0.80       792
           1       0.50      0.17      0.25         6
           2       0.36      0.18      0.24        28
           3       0.00      0.00      0.00         6
           4       0.57      0.39      0.47        51
           5       0.52      0.38      0.44        29
           6       0.57      0.33      0.42        24
           7       0.00      0.00      0.00         8
           8       0.00      0.00      0.00        17
           9       0.94      0.94      0.94        16
          10       0.50      0.33      0.40        18
          11       0.25      0.07      0.11        43
          12       0.45      0.35      0.40        48
          13       0.00      0.00      0.00         7
          14       0.00      0.00      0.00         6
          15       0.50      0.17      0.25         6
          16       0.00      0.00      0.00         5
          17       0.64      0.80      0.71        56
          18       0.20      0.04      0.07        23
          19       0.00      0.00      0.00        11
          20       0.00      0.00      0.00         4
          21       0.00      0.00      0.00         9
          22       0.50      0.16      0.24        19
          23       0.50      0.12      0.20        40
          24       1.00      0.29      0.44         7
          25       0.00      0.00      0.00        14
          26       0.00      0.00      0.00         1
          27       0.50      0.33      0.40        21
          28       0.00      0.00      0.00        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       1.00      0.25      0.40         4
          33       1.00      0.10      0.18        20
          34       1.00      0.22      0.36         9
          35       1.00      0.25      0.40         8
          36       0.33      0.14      0.20         7
          37       0.00      0.00      0.00         1
          38       0.50      0.33      0.40         3
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       0.00      0.00      0.00         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.33      0.04      0.07        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         1
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.43      0.17      0.24        36
          57       1.00      0.20      0.33         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       0.00      0.00      0.00         1
          63       0.57      0.29      0.38        14
          66       0.54      0.92      0.68       132
          67       0.50      0.10      0.17        50

    accuracy                           0.63      1687
   macro avg       0.30      0.16      0.19      1687
weighted avg       0.55      0.63      0.55      1687

F1 score:  0.7145924962541153
Time taken: 1.754265s to run the model

Random Forest

df_RF_Results = pd.DataFrame(columns=['Model','Train Accuracy','Test Accuracy','F1 Score','Time Taken (sec)'])

model_name = 'Random Forest'
df_temp = run_classification(model_name, RandomForestClassifier(n_estimators=100), X_train, X_test, y_train, y_test)
df_RF_Results = pd.concat([df_RF_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)] = df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=Non...
                 RandomForestClassifier(bootstrap=True, ccp_alpha=0.0,
                                        class_weight=None, criterion='gini',
                                        max_depth=None, max_features='auto',
                                        max_leaf_nodes=None, max_samples=None,
                                        min_impurity_decrease=0.0,
                                        min_impurity_split=None,
                                        min_samples_leaf=1, min_samples_split=2,
                                        min_weight_fraction_leaf=0.0,
                                        n_estimators=100, n_jobs=None,
                                        oob_score=False, random_state=None,
                                        verbose=0, warm_start=False))],
         verbose=False)
================================================================================
Training accuracy:  0.9349051008303677
Testing accuracy:  0.6502667457024304
================================================================================
Confusion matrix:
  [[781   0   0 ...   2   0   0]
 [  1   2   0 ...   0   3   0]
 [  8   0   5 ...   0  12   0]
 ...
 [ 10   0   0 ...   4   0   0]
 [  2   0   0 ...   0 125   0]
 [ 11   0   0 ...   0  34   3]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.66      0.99      0.79       792
           1       0.67      0.33      0.44         6
           2       1.00      0.18      0.30        28
           3       0.00      0.00      0.00         6
           4       0.71      0.33      0.45        51
           5       0.47      0.48      0.47        29
           6       0.67      0.33      0.44        24
           7       0.00      0.00      0.00         8
           8       1.00      0.06      0.11        17
           9       1.00      0.94      0.97        16
          10       0.44      0.22      0.30        18
          11       0.31      0.09      0.14        43
          12       0.76      0.33      0.46        48
          13       0.00      0.00      0.00         7
          14       1.00      0.17      0.29         6
          15       0.00      0.00      0.00         6
          16       0.00      0.00      0.00         5
          17       0.73      0.79      0.76        56
          18       0.80      0.35      0.48        23
          19       0.00      0.00      0.00        11
          20       0.00      0.00      0.00         4
          21       0.00      0.00      0.00         9
          22       0.67      0.42      0.52        19
          23       1.00      0.20      0.33        40
          24       1.00      0.29      0.44         7
          25       1.00      0.21      0.35        14
          26       0.00      0.00      0.00         1
          27       0.56      0.24      0.33        21
          28       1.00      0.17      0.29        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       1.00      0.25      0.40         4
          33       0.67      0.10      0.17        20
          34       0.00      0.00      0.00         9
          35       1.00      0.12      0.22         8
          36       0.00      0.00      0.00         7
          37       0.00      0.00      0.00         1
          38       0.00      0.00      0.00         3
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       0.00      0.00      0.00         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.00      0.00      0.00        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         1
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.73      0.31      0.43        36
          57       0.00      0.00      0.00         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       1.00      1.00      1.00         1
          61       0.00      0.00      0.00         1
          63       0.67      0.29      0.40        14
          66       0.54      0.95      0.69       132
          67       0.50      0.06      0.11        50

    accuracy                           0.65      1687
   macro avg       0.36      0.18      0.21      1687
weighted avg       0.61      0.65      0.57      1687

F1 score:  0.7305030484294792
Time taken: 10.000938s to run the model

model_name = 'Random Forest with b-weights'
df_temp = run_classification(model_name, RandomForestClassifier(n_estimators=100, class_weight = 'balanced'), X_train, X_test, y_train, y_test)
df_RF_Results = pd.concat([df_RF_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)] = df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=Non...
                 RandomForestClassifier(bootstrap=True, ccp_alpha=0.0,
                                        class_weight='balanced',
                                        criterion='gini', max_depth=None,
                                        max_features='auto',
                                        max_leaf_nodes=None, max_samples=None,
                                        min_impurity_decrease=0.0,
                                        min_impurity_split=None,
                                        min_samples_leaf=1, min_samples_split=2,
                                        min_weight_fraction_leaf=0.0,
                                        n_estimators=100, n_jobs=None,
                                        oob_score=False, random_state=None,
                                        verbose=0, warm_start=False))],
         verbose=False)
================================================================================
Training accuracy:  0.8913107947805456
Testing accuracy:  0.5909899229401304
================================================================================
Confusion matrix:
  [[762   0   0 ...   9   0   0]
 [  2   1   0 ...   0   0   0]
 [ 10   0   4 ...   0   0   0]
 ...
 [  0   0   0 ...   0   0   0]
 [  2   0   0 ...   0  49   0]
 [ 12   0   0 ...   0   0   3]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.63      0.96      0.76       792
           1       1.00      0.17      0.29         6
           2       1.00      0.14      0.25        28
           3       0.00      0.00      0.00         6
           4       0.71      0.24      0.35        51
           5       0.38      0.17      0.24        29
           6       0.69      0.38      0.49        24
           7       0.00      0.00      0.00         8
           8       1.00      0.12      0.21        17
           9       0.94      0.94      0.94        16
          10       0.60      0.17      0.26        18
          11       0.33      0.07      0.12        43
          12       0.64      0.29      0.40        48
          13       0.00      0.00      0.00         7
          14       1.00      0.17      0.29         6
          15       0.00      0.00      0.00         6
          16       0.50      0.20      0.29         5
          17       0.78      0.75      0.76        56
          18       0.89      0.35      0.50        23
          19       0.00      0.00      0.00        11
          20       0.00      0.00      0.00         4
          21       0.00      0.00      0.00         9
          22       0.73      0.42      0.53        19
          23       1.00      0.07      0.14        40
          24       1.00      0.14      0.25         7
          25       1.00      0.21      0.35        14
          26       0.00      0.00      0.00         1
          27       0.44      0.19      0.27        21
          28       0.17      0.08      0.11        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       1.00      0.25      0.40         4
          33       1.00      0.15      0.26        20
          34       0.00      0.00      0.00         9
          35       1.00      0.25      0.40         8
          36       0.00      0.00      0.00         7
          37       0.00      0.00      0.00         1
          38       0.00      0.00      0.00         3
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       0.11      0.60      0.19         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.12      0.65      0.20        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         1
          53       0.00      0.00      0.00         1
          54       0.00      0.00      0.00         1
          55       0.54      0.19      0.29        36
          57       0.67      0.40      0.50         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       1.00      1.00      1.00         1
          61       0.00      0.00      0.00         1
          63       0.78      0.50      0.61        14
          65       0.00      0.00      0.00         0
          66       0.88      0.37      0.52       132
          67       1.00      0.06      0.11        50

    accuracy                           0.59      1687
   macro avg       0.37      0.17      0.19      1687
weighted avg       0.64      0.59      0.53      1687

F1 score:  0.6512691612865644
Time taken: 10.896074s to run the model

model_name = 'Random Forest without word count features'
df_temp = run_classification(model_name, RandomForestClassifier(n_estimators=100), X_train_v, X_test_v, y_train_v, y_test_v, arch_name= None, pipelineRequired=False)
df_RF_Results = pd.concat([df_RF_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)] = df_temp

start
Estimator: RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,
                       criterion='gini', max_depth=None, max_features='auto',
                       max_leaf_nodes=None, max_samples=None,
                       min_impurity_decrease=0.0, min_impurity_split=None,
                       min_samples_leaf=1, min_samples_split=2,
                       min_weight_fraction_leaf=0.0, n_estimators=100,
                       n_jobs=None, oob_score=False, random_state=None,
                       verbose=0, warm_start=False)
================================================================================
Training accuracy:  0.9306049822064056
Testing accuracy:  0.6597510373443983
================================================================================
Confusion matrix:
  [[779   0   0 ...   2   0   0]
 [  0   0   0 ...   0   3   0]
 [ 16   0   1 ...   0  11   0]
 ...
 [  3   0   0 ...  10   0   0]
 [  5   0   0 ...   0 137   1]
 [  6   0   0 ...   0  28   3]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.69      0.97      0.80       803
           1       0.00      0.00      0.00         3
           2       0.50      0.03      0.06        31
           3       0.00      0.00      0.00         9
           4       0.56      0.38      0.45        39
           5       0.56      0.36      0.43        28
           6       0.56      0.25      0.34        20
           7       1.00      0.22      0.36         9
           8       0.75      0.20      0.32        15
           9       1.00      0.82      0.90        17
          10       0.36      0.21      0.27        19
          11       0.46      0.11      0.17        56
          12       0.75      0.44      0.55        48
          13       0.00      0.00      0.00         7
          14       0.00      0.00      0.00         7
          15       0.00      0.00      0.00         4
          16       0.33      0.25      0.29         4
          17       0.77      0.85      0.81        47
          18       0.71      0.31      0.43        32
          19       0.00      0.00      0.00        12
          20       0.00      0.00      0.00         1
          21       1.00      0.09      0.17        11
          22       0.57      0.50      0.53        16
          23       0.71      0.13      0.22        38
          24       1.00      0.12      0.22         8
          25       0.67      0.13      0.22        15
          26       0.00      0.00      0.00         1
          27       0.43      0.14      0.21        22
          28       0.67      0.15      0.25        13
          29       0.00      0.00      0.00         2
          30       1.00      0.17      0.29         6
          32       1.00      0.50      0.67         2
          33       1.00      0.20      0.33        15
          34       0.00      0.00      0.00        11
          35       1.00      0.29      0.44         7
          36       0.00      0.00      0.00         3
          37       0.00      0.00      0.00         1
          38       1.00      0.20      0.33         5
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       1.00      0.29      0.44         7
          42       0.26      1.00      0.41         8
          43       0.00      0.00      0.00         3
          44       0.50      0.05      0.08        22
          45       0.00      0.00      0.00         2
          46       0.00      0.00      0.00         1
          47       0.00      0.00      0.00         1
          48       0.00      0.00      0.00         2
          50       1.00      1.00      1.00         2
          51       0.00      0.00      0.00         2
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.47      0.24      0.32        29
          57       1.00      0.14      0.25         7
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       0.00      0.00      0.00         1
          63       0.83      0.77      0.80        13
          66       0.57      0.93      0.71       148
          67       0.75      0.08      0.14        38

    accuracy                           0.66      1687
   macro avg       0.43      0.22      0.25      1687
weighted avg       0.63      0.66      0.59      1687

F1 score:  0.7330616613250273
Time taken: 38.389570s to run the model

model_name = 'Random Forest with word count features'
df_temp = run_classification(model_name, RandomForestClassifier(n_estimators=100), X_train_vf, X_test_vf, y_train_vf, y_test_vf, arch_name= None, pipelineRequired=False)
df_RF_Results = pd.concat([df_RF_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)] = df_temp

start
Estimator: RandomForestClassifier(bootstrap=True, ccp_alpha=0.0, class_weight=None,
                       criterion='gini', max_depth=None, max_features='auto',
                       max_leaf_nodes=None, max_samples=None,
                       min_impurity_decrease=0.0, min_impurity_split=None,
                       min_samples_leaf=1, min_samples_split=2,
                       min_weight_fraction_leaf=0.0, n_estimators=100,
                       n_jobs=None, oob_score=False, random_state=None,
                       verbose=0, warm_start=False)
================================================================================
Training accuracy:  0.9344602609727165
Testing accuracy:  0.6532305868405454
================================================================================
Confusion matrix:
  [[781   0   0 ...   3   0   0]
 [  0   0   0 ...   0   3   0]
 [ 17   0   1 ...   0  11   0]
 ...
 [  3   0   0 ...  10   0   0]
 [  5   0   0 ...   0 137   1]
 [  7   0   0 ...   0  28   3]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.67      0.97      0.79       803
           1       0.00      0.00      0.00         3
           2       1.00      0.03      0.06        31
           3       0.00      0.00      0.00         9
           4       0.57      0.33      0.42        39
           5       0.53      0.32      0.40        28
           6       0.62      0.25      0.36        20
           7       1.00      0.22      0.36         9
           8       0.83      0.33      0.48        15
           9       0.93      0.82      0.87        17
          10       0.50      0.26      0.34        19
          11       0.33      0.05      0.09        56
          12       0.71      0.42      0.53        48
          13       0.00      0.00      0.00         7
          14       0.00      0.00      0.00         7
          15       0.00      0.00      0.00         4
          16       0.50      0.25      0.33         4
          17       0.88      0.81      0.84        47
          18       0.73      0.34      0.47        32
          19       0.00      0.00      0.00        12
          20       0.00      0.00      0.00         1
          21       1.00      0.09      0.17        11
          22       0.75      0.38      0.50        16
          23       0.33      0.03      0.05        38
          24       0.25      0.38      0.30         8
          25       0.43      0.20      0.27        15
          26       0.00      0.00      0.00         1
          27       0.83      0.23      0.36        22
          28       0.67      0.15      0.25        13
          29       0.00      0.00      0.00         2
          30       0.00      0.00      0.00         6
          32       1.00      0.50      0.67         2
          33       0.75      0.20      0.32        15
          34       0.00      0.00      0.00        11
          35       1.00      0.29      0.44         7
          36       0.00      0.00      0.00         3
          37       0.00      0.00      0.00         1
          38       1.00      0.20      0.33         5
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       1.00      0.14      0.25         7
          42       0.45      0.62      0.53         8
          43       0.00      0.00      0.00         3
          44       0.50      0.05      0.08        22
          45       0.00      0.00      0.00         2
          46       0.00      0.00      0.00         1
          47       0.00      0.00      0.00         1
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         2
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.58      0.24      0.34        29
          57       1.00      0.14      0.25         7
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       0.00      0.00      0.00         1
          63       0.71      0.77      0.74        13
          66       0.57      0.93      0.71       148
          67       0.75      0.08      0.14        38

    accuracy                           0.65      1687
   macro avg       0.40      0.20      0.23      1687
weighted avg       0.62      0.65      0.57      1687

F1 score:  0.7316106021087403
Time taken: 34.854284s to run the model

df_RF_Results.sort_values(by='Test Accuracy', ascending = False)

	Model	Train Accuracy	Test Accuracy	F1 Score	Time Taken (sec)
0	Random Forest without word count features	0.930605	0.659751	0.733062	38.483404
0	Random Forest with word count features	0.934460	0.653231	0.731611	34.940396
0	Random Forest	0.934905	0.650267	0.730503	10.093235
0	Random Forest with b-weights	0.891311	0.590990	0.651269	10.988256

df_RF_Results.plot(kind='barh',x='Model', figsize=[15,10])

<matplotlib.axes._subplots.AxesSubplot at 0x7f70981e03d0>

Decision trees

df_DT_Results = pd.DataFrame(columns=['Model','Train Accuracy','Test Accuracy','F1 Score','Time Taken (sec)'])

model_name = 'Decision trees'
df_temp = run_classification(model_name, DecisionTreeClassifier(), X_train, X_test, y_train, y_test)
df_DT_Results = pd.concat([df_DT_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)] = df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=Non...
                                  sublinear_tf=False, use_idf=True)),
                ('clf',
                 DecisionTreeClassifier(ccp_alpha=0.0, class_weight=None,
                                        criterion='gini', max_depth=None,
                                        max_features=None, max_leaf_nodes=None,
                                        min_impurity_decrease=0.0,
                                        min_impurity_split=None,
                                        min_samples_leaf=1, min_samples_split=2,
                                        min_weight_fraction_leaf=0.0,
                                        presort='deprecated', random_state=None,
                                        splitter='best'))],
         verbose=False)
================================================================================
Training accuracy:  0.9349051008303677
Testing accuracy:  0.5909899229401304
================================================================================
Confusion matrix:
  [[665   0   3 ...   4   1   2]
 [  0   1   0 ...   0   3   0]
 [  5   0   5 ...   0  12   0]
 ...
 [  5   0   0 ...   8   0   0]
 [  1   0   0 ...   0 114   0]
 [  5   0   0 ...   0  34   7]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.75      0.84      0.79       792
           1       0.20      0.17      0.18         6
           2       0.36      0.18      0.24        28
           3       0.25      0.17      0.20         6
           4       0.49      0.39      0.43        51
           5       0.38      0.45      0.41        29
           6       0.29      0.25      0.27        24
           7       0.57      0.50      0.53         8
           8       0.47      0.41      0.44        17
           9       0.88      0.94      0.91        16
          10       0.25      0.22      0.24        18
          11       0.29      0.23      0.26        43
          12       0.44      0.33      0.38        48
          13       0.00      0.00      0.00         7
          14       0.40      0.33      0.36         6
          15       0.17      0.17      0.17         6
          16       0.14      0.20      0.17         5
          17       0.66      0.73      0.69        56
          18       0.52      0.48      0.50        23
          19       0.00      0.00      0.00        11
          20       0.00      0.00      0.00         4
          21       0.00      0.00      0.00         9
          22       0.32      0.42      0.36        19
          23       0.23      0.17      0.20        40
          24       0.60      0.43      0.50         7
          25       0.17      0.07      0.10        14
          26       0.00      0.00      0.00         1
          27       0.19      0.19      0.19        21
          28       0.43      0.25      0.32        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       1.00      0.25      0.40         4
          33       0.11      0.10      0.11        20
          34       0.00      0.00      0.00         9
          35       0.44      0.50      0.47         8
          36       0.00      0.00      0.00         7
          37       0.00      0.00      0.00         1
          38       0.00      0.00      0.00         3
          39       0.20      0.14      0.17         7
          40       0.00      0.00      0.00         1
          41       0.50      0.20      0.29         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.00      0.00      0.00        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.50      0.50      0.50         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         1
          53       0.00      0.00      0.00         1
          54       0.00      0.00      0.00         1
          55       0.37      0.19      0.25        36
          57       0.33      0.20      0.25         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       1.00      1.00      1.00         1
          61       0.00      0.00      0.00         1
          63       0.53      0.57      0.55        14
          66       0.53      0.86      0.65       132
          67       0.50      0.14      0.22        50

    accuracy                           0.59      1687
   macro avg       0.25      0.21      0.22      1687
weighted avg       0.55      0.59      0.56      1687

F1 score:  0.6207338067561049
Time taken: 1.660094s to run the model

model_name = 'Decision trees with b-weights'
df_temp = run_classification(model_name, DecisionTreeClassifier(class_weight='balanced'), X_train, X_test, y_train, y_test)
df_DT_Results = pd.concat([df_DT_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)] = df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=Non...
                                  sublinear_tf=False, use_idf=True)),
                ('clf',
                 DecisionTreeClassifier(ccp_alpha=0.0, class_weight='balanced',
                                        criterion='gini', max_depth=None,
                                        max_features=None, max_leaf_nodes=None,
                                        min_impurity_decrease=0.0,
                                        min_impurity_split=None,
                                        min_samples_leaf=1, min_samples_split=2,
                                        min_weight_fraction_leaf=0.0,
                                        presort='deprecated', random_state=None,
                                        splitter='best'))],
         verbose=False)
================================================================================
Training accuracy:  0.8827105575326216
Testing accuracy:  0.4872554831061055
================================================================================
Confusion matrix:
  [[547   0   8 ...   9   3   3]
 [  1   1   0 ...   0   0   0]
 [  1   0   3 ...   0   0   1]
 ...
 [  0   0   0 ...   0   0   0]
 [  2   0   0 ...   0  38   0]
 [  2   0   1 ...   0   0   9]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.80      0.69      0.74       792
           1       0.50      0.17      0.25         6
           2       0.13      0.11      0.12        28
           3       0.33      0.33      0.33         6
           4       0.28      0.24      0.26        51
           5       0.33      0.31      0.32        29
           6       0.17      0.21      0.19        24
           7       0.10      0.12      0.11         8
           8       0.30      0.35      0.32        17
           9       0.88      0.94      0.91        16
          10       0.18      0.22      0.20        18
          11       0.25      0.26      0.25        43
          12       0.34      0.42      0.37        48
          13       0.00      0.00      0.00         7
          14       0.40      0.33      0.36         6
          15       0.20      0.33      0.25         6
          16       0.20      0.20      0.20         5
          17       0.66      0.68      0.67        56
          18       0.34      0.43      0.38        23
          19       0.04      0.09      0.06        11
          20       0.00      0.00      0.00         4
          21       0.17      0.11      0.13         9
          22       0.36      0.47      0.41        19
          23       0.16      0.12      0.14        40
          24       0.17      1.00      0.29         7
          25       0.27      0.29      0.28        14
          26       0.00      0.00      0.00         1
          27       0.19      0.19      0.19        21
          28       0.08      0.08      0.08        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       0.25      0.25      0.25         4
          33       0.16      0.25      0.20        20
          34       0.14      0.11      0.12         9
          35       0.46      0.75      0.57         8
          36       0.17      0.14      0.15         7
          37       0.00      0.00      0.00         1
          38       0.33      0.33      0.33         3
          39       0.00      0.00      0.00         7
          40       0.20      1.00      0.33         1
          41       0.08      0.40      0.13         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.12      0.65      0.20        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       1.00      0.50      0.67         2
          50       0.50      0.50      0.50         2
          51       0.00      0.00      0.00         1
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.38      0.14      0.20        36
          57       0.29      0.40      0.33         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.25      1.00      0.40         1
          61       0.33      1.00      0.50         1
          63       0.35      0.50      0.41        14
          64       0.00      0.00      0.00         0
          65       0.00      0.00      0.00         0
          66       0.75      0.29      0.42       132
          67       0.45      0.18      0.26        50

    accuracy                           0.49      1687
   macro avg       0.23      0.28      0.23      1687
weighted avg       0.57      0.49      0.51      1687

F1 score:  0.4671061730791086
Time taken: 2.412444s to run the model

model_name = 'Decision trees without word count features'
df_temp = run_classification(model_name, DecisionTreeClassifier(), X_train_v, X_test_v, y_train_v, y_test_v, arch_name= None, pipelineRequired=False)
df_DT_Results = pd.concat([df_DT_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)] = df_temp

start
Estimator: DecisionTreeClassifier(ccp_alpha=0.0, class_weight=None, criterion='gini',
                       max_depth=None, max_features=None, max_leaf_nodes=None,
                       min_impurity_decrease=0.0, min_impurity_split=None,
                       min_samples_leaf=1, min_samples_split=2,
                       min_weight_fraction_leaf=0.0, presort='deprecated',
                       random_state=None, splitter='best')
================================================================================
Training accuracy:  0.9307532621589562
Testing accuracy:  0.6206283343212804
================================================================================
Confusion matrix:
  [[685   0   0 ...   1   1   3]
 [  0   0   0 ...   0   3   0]
 [  6   0   4 ...   0  11   0]
 ...
 [  3   0   0 ...   9   0   0]
 [  3   0   0 ...   0 129   1]
 [  1   0   0 ...   0  28   6]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.78      0.85      0.81       803
           1       0.00      0.00      0.00         3
           2       0.44      0.13      0.20        31
           3       0.00      0.00      0.00         9
           4       0.49      0.51      0.50        39
           5       0.18      0.18      0.18        28
           6       0.29      0.20      0.24        20
           7       0.60      0.33      0.43         9
           8       0.35      0.40      0.38        15
           9       1.00      1.00      1.00        17
          10       0.23      0.16      0.19        19
          11       0.39      0.21      0.28        56
          12       0.42      0.50      0.46        48
          13       0.00      0.00      0.00         7
          14       0.33      0.14      0.20         7
          15       0.50      0.25      0.33         4
          16       0.20      0.25      0.22         4
          17       0.83      0.81      0.82        47
          18       0.52      0.44      0.47        32
          19       0.00      0.00      0.00        12
          20       0.00      0.00      0.00         1
          21       0.12      0.09      0.11        11
          22       0.38      0.50      0.43        16
          23       0.21      0.18      0.20        38
          24       1.00      0.12      0.22         8
          25       0.50      0.27      0.35        15
          26       0.00      0.00      0.00         1
          27       0.40      0.27      0.32        22
          28       0.67      0.31      0.42        13
          29       0.00      0.00      0.00         2
          30       0.00      0.00      0.00         6
          32       1.00      0.50      0.67         2
          33       0.29      0.33      0.31        15
          34       0.43      0.27      0.33        11
          35       1.00      0.57      0.73         7
          36       0.00      0.00      0.00         3
          37       0.00      0.00      0.00         1
          38       0.50      0.20      0.29         5
          39       0.33      0.29      0.31         7
          40       0.00      0.00      0.00         1
          41       1.00      0.14      0.25         7
          42       0.15      1.00      0.25         8
          43       0.00      0.00      0.00         3
          44       0.20      0.05      0.07        22
          45       0.00      0.00      0.00         2
          46       0.00      0.00      0.00         1
          47       0.00      0.00      0.00         1
          48       0.00      0.00      0.00         2
          49       0.00      0.00      0.00         0
          50       0.33      0.50      0.40         2
          51       0.00      0.00      0.00         2
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.36      0.17      0.23        29
          57       0.50      0.14      0.22         7
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       0.00      0.00      0.00         1
          63       0.82      0.69      0.75        13
          66       0.56      0.87      0.68       148
          67       0.55      0.16      0.24        38

    accuracy                           0.62      1687
   macro avg       0.32      0.24      0.25      1687
weighted avg       0.60      0.62      0.59      1687

F1 score:  0.6470369922976751
Time taken: 12.421041s to run the model

model_name = 'Decision trees with word count features'
df_temp = run_classification(model_name, DecisionTreeClassifier(), X_train_vf, X_test_vf, y_train_vf, y_test_vf, arch_name= None, pipelineRequired=False)
df_DT_Results = pd.concat([df_DT_Results, df_temp])
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)] = df_temp

start
Estimator: DecisionTreeClassifier(ccp_alpha=0.0, class_weight=None, criterion='gini',
                       max_depth=None, max_features=None, max_leaf_nodes=None,
                       min_impurity_decrease=0.0, min_impurity_split=None,
                       min_samples_leaf=1, min_samples_split=2,
                       min_weight_fraction_leaf=0.0, presort='deprecated',
                       random_state=None, splitter='best')
================================================================================
Training accuracy:  0.9344602609727165
Testing accuracy:  0.5815056312981625
================================================================================
Confusion matrix:
  [[653   0   1 ...   3   0   3]
 [  0   0   0 ...   0   2   0]
 [  8   0   2 ...   0  11   0]
 ...
 [  1   0   0 ...  10   0   0]
 [  3   0   0 ...   0 128   1]
 [  3   0   0 ...   0  27   6]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.74      0.81      0.78       803
           1       0.00      0.00      0.00         3
           2       0.18      0.06      0.10        31
           3       0.00      0.00      0.00         9
           4       0.45      0.44      0.44        39
           5       0.29      0.25      0.27        28
           6       0.19      0.20      0.20        20
           7       0.30      0.33      0.32         9
           8       0.40      0.27      0.32        15
           9       0.87      0.76      0.81        17
          10       0.23      0.16      0.19        19
          11       0.36      0.32      0.34        56
          12       0.37      0.42      0.39        48
          13       0.17      0.14      0.15         7
          14       0.50      0.14      0.22         7
          15       0.00      0.00      0.00         4
          16       0.50      0.50      0.50         4
          17       0.88      0.81      0.84        47
          18       0.61      0.34      0.44        32
          19       0.00      0.00      0.00        12
          20       0.00      0.00      0.00         1
          21       0.00      0.00      0.00        11
          22       0.29      0.31      0.30        16
          23       0.15      0.08      0.10        38
          24       0.10      0.12      0.11         8
          25       0.12      0.13      0.13        15
          26       0.00      0.00      0.00         1
          27       0.20      0.18      0.19        22
          28       0.10      0.08      0.09        13
          29       0.00      0.00      0.00         2
          30       0.00      0.00      0.00         6
          31       0.00      0.00      0.00         0
          32       0.50      0.50      0.50         2
          33       0.24      0.33      0.28        15
          34       0.00      0.00      0.00        11
          35       0.80      0.57      0.67         7
          36       0.00      0.00      0.00         3
          37       0.00      0.00      0.00         1
          38       0.50      0.20      0.29         5
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       1.00      0.14      0.25         7
          42       0.21      0.38      0.27         8
          43       0.00      0.00      0.00         3
          44       0.25      0.05      0.08        22
          45       0.00      0.00      0.00         2
          46       1.00      1.00      1.00         1
          47       0.00      0.00      0.00         1
          48       0.00      0.00      0.00         2
          50       0.33      0.50      0.40         2
          51       0.00      0.00      0.00         2
          53       1.00      1.00      1.00         1
          54       0.00      0.00      0.00         1
          55       0.31      0.14      0.19        29
          57       0.25      0.14      0.18         7
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       0.00      0.00      0.00         1
          63       0.71      0.77      0.74        13
          66       0.57      0.86      0.69       148
          67       0.38      0.16      0.22        38

    accuracy                           0.58      1687
   macro avg       0.26      0.22      0.23      1687
weighted avg       0.55      0.58      0.56      1687

F1 score:  0.605928666774809
Time taken: 6.981524s to run the model

df_DT_Results.sort_values(by='Test Accuracy', ascending = False)

	Model	Train Accuracy	Test Accuracy	F1 Score	Time Taken (sec)
0	Decision trees without word count features	0.930753	0.620628	0.647037	12.503277
0	Decision trees	0.934905	0.590990	0.620734	1.742803
0	Decision trees with word count features	0.934460	0.581506	0.605929	7.063638
0	Decision trees with b-weights	0.882711	0.487255	0.467106	2.505630

df_DT_Results.plot(kind='barh',x='Model', figsize=[15,10])

<matplotlib.axes._subplots.AxesSubplot at 0x7f70987461d0>

xgboost

model_name = 'xgboost'
df_temp = run_classification(model_name, xgb.XGBClassifier(max_depth=7, n_estimators=200, colsample_bytree=0.8, 
                        subsample=0.8, nthread=10, learning_rate=0.1), X_train, X_test, y_train, y_test)
if model_name not in df_resLog.values :
  df_resLog = pd.concat([df_resLog, df_temp])
else:
  df_resLog.loc[(df_resLog.Model == model_name)] = df_temp

start
Estimator: Pipeline(memory=None,
         steps=[('vect',
                 CountVectorizer(analyzer='word', binary=False,
                                 decode_error='strict',
                                 dtype=<class 'numpy.int64'>, encoding='utf-8',
                                 input='content', lowercase=True, max_df=1.0,
                                 max_features=None, min_df=1,
                                 ngram_range=(1, 1), preprocessor=None,
                                 stop_words=None, strip_accents=None,
                                 token_pattern='(?u)\\b\\w\\w+\\b',
                                 tokenizer=None, vocabulary=Non...
                 XGBClassifier(base_score=0.5, booster='gbtree',
                               colsample_bylevel=1, colsample_bynode=1,
                               colsample_bytree=0.8, gamma=0, learning_rate=0.1,
                               max_delta_step=0, max_depth=7,
                               min_child_weight=1, missing=None,
                               n_estimators=200, n_jobs=1, nthread=10,
                               objective='multi:softprob', random_state=0,
                               reg_alpha=0, reg_lambda=1, scale_pos_weight=1,
                               seed=None, silent=None, subsample=0.8,
                               verbosity=1))],
         verbose=False)
================================================================================
Training accuracy:  0.8871589561091341
Testing accuracy:  0.6395969176052163
================================================================================
Confusion matrix:
  [[744   0   1 ...   3   0   2]
 [  0   1   0 ...   0   3   0]
 [  5   0   7 ...   0  12   0]
 ...
 [  4   0   0 ...   9   0   0]
 [  2   0   0 ...   0 120   0]
 [  7   0   0 ...   0  34   5]]
================================================================================
Classification report:
                precision    recall  f1-score   support

           0       0.71      0.94      0.81       792
           1       0.25      0.17      0.20         6
           2       0.70      0.25      0.37        28
           3       0.33      0.17      0.22         6
           4       0.53      0.45      0.49        51
           5       0.45      0.52      0.48        29
           6       0.50      0.46      0.48        24
           7       0.00      0.00      0.00         8
           8       0.69      0.53      0.60        17
           9       0.93      0.88      0.90        16
          10       0.36      0.22      0.28        18
          11       0.42      0.19      0.26        43
          12       0.46      0.33      0.39        48
          13       0.00      0.00      0.00         7
          14       0.67      0.33      0.44         6
          15       0.67      0.67      0.67         6
          16       0.20      0.20      0.20         5
          17       0.74      0.66      0.70        56
          18       0.69      0.48      0.56        23
          19       0.00      0.00      0.00        11
          20       0.00      0.00      0.00         4
          21       0.00      0.00      0.00         9
          22       0.53      0.53      0.53        19
          23       0.31      0.12      0.18        40
          24       0.00      0.00      0.00         7
          25       0.75      0.21      0.33        14
          26       0.00      0.00      0.00         1
          27       0.36      0.24      0.29        21
          28       0.67      0.17      0.27        12
          29       0.00      0.00      0.00         3
          30       0.00      0.00      0.00         3
          31       0.00      0.00      0.00         1
          32       1.00      0.25      0.40         4
          33       0.20      0.05      0.08        20
          34       0.00      0.00      0.00         9
          35       0.67      0.50      0.57         8
          36       0.00      0.00      0.00         7
          37       0.00      0.00      0.00         1
          38       0.00      0.00      0.00         3
          39       0.00      0.00      0.00         7
          40       0.00      0.00      0.00         1
          41       0.00      0.00      0.00         5
          42       0.00      0.00      0.00         5
          43       0.00      0.00      0.00         1
          44       0.00      0.00      0.00        26
          45       0.00      0.00      0.00         3
          46       0.00      0.00      0.00         2
          47       0.00      0.00      0.00         2
          48       0.00      0.00      0.00         2
          50       0.00      0.00      0.00         2
          51       0.00      0.00      0.00         1
          53       0.00      0.00      0.00         1
          54       0.00      0.00      0.00         1
          55       0.55      0.17      0.26        36
          57       0.00      0.00      0.00         5
          58       0.00      0.00      0.00         1
          59       0.00      0.00      0.00         2
          60       0.00      0.00      0.00         1
          61       0.00      0.00      0.00         1
          63       0.64      0.64      0.64        14
          66       0.54      0.91      0.67       132
          67       0.56      0.10      0.17        50

    accuracy                           0.64      1687
   macro avg       0.26      0.18      0.20      1687
weighted avg       0.57      0.64      0.58      1687

F1 score:  0.6993235931011635
Time taken: 293.300965s to run the model

df_resLog = df_resLog.sort_values(by='Test Accuracy', ascending = False)
df_resLog

	Model	Train Accuracy	Test Accuracy	F1 Score	Time Taken (sec)
0	Random Forest without word count features	0.930605	0.659751	0.733062	38.483404
0	Stochastic Gradient Descent	0.762011	0.654416	0.733147	5.142270
0	Random Forest with word count features	0.934460	0.653231	0.731611	34.940396
0	Random Forest	0.934905	0.650267	0.730503	10.093235
0	Multinomial Logistic Regression with word coun...	0.929270	0.646117	0.660112	119.157214
0	Multinomial Logistic Regression without word c...	0.929715	0.641968	0.654509	114.525980
0	xgboost	0.887159	0.639597	0.699324	293.403598
0	K Nearest neighbor	0.693357	0.633669	0.714592	1.840264
0	Multinomial Logistic Regression	0.663553	0.621814	0.722842	18.808882
0	Decision trees without word count features	0.930753	0.620628	0.647037	12.503277
0	SVM-RBF Kernal	0.793743	0.620036	0.723278	126.391750
0	Random Forest with b-weights	0.891311	0.590990	0.651269	10.988256
0	Decision trees	0.934905	0.590990	0.620734	1.742803
0	Decision trees with word count features	0.934460	0.581506	0.605929	7.063638
0	Multinomial Naive Bayes	0.568209	0.554831	0.701626	0.372570
0	Stochastic Gradient Descent with b-weights	0.680308	0.491998	0.437865	5.021881
0	Decision trees with b-weights	0.882711	0.487255	0.467106	2.505630
0	SVM-RBF Kernal with b-weights	0.631673	0.451689	0.404652	209.575290
0	Multinomial Logistic Regression with b-weights	0.502966	0.378779	0.342945	11.690096

df_resLog.plot(kind='barh',x='Model', figsize=[20,15])

<matplotlib.axes._subplots.AxesSubplot at 0x7f7082100050>

Observations:

• Random forest, Xgboost and Decision Tree seems to be overfitting due to high difference in Training and test accuracy

• F1 score is higher than accuracy because it is calculated for the target group present in y_pred (lesser than the y_train ) Class Imbalance

• The best performing model “ Stochastic Gradient Descent” performance reduces drastically from 65.44% to 49.19% once the parameter class_weight = ‘balanced'

• K NN, Xgboost and Random Forest seem to handle imbalance better than SGD.

• Meanwhile, SGD, SVM and Naïve Bayes might perform better if the class imbalance is taken care

• Vectorized input with 'Max features' = 12270 is quiet a high dimension for SGD, MultinomialNB, KNN. It takes a long time to converge. Max features less than 1000 does not give a good performance.

• The vectorized input with Max features = 12270 improved the performance for Multinomial Logistic Regression, Random Forest & Decision Tree - Handles high dimension better

• Addition word count features to the input dataframe does not add much value to the perfomance.

• List item

• SVM (RBF Kernel) and Xgboost seems to take a lot of time for execution compared to others even for lesser dimensional input. Thus avoided modelling with vectorized input.

Deep Learning Models

Neural Network

tfidf = TfidfVectorizer(max_features=250, analyzer = 'word', min_df=2, max_df=0.95, ngram_range=(1, 2))
tckt_tfidf = tfidf.fit_transform(ticket_df['tokenized_desc'])

x = tckt_tfidf.toarray()
y=ticket_df["LabelEncodings"]

## Splitting the data into train and test
from sklearn.model_selection import train_test_split
X_train_n,X_test_n,y_train_n,y_test_n=train_test_split(x,y,test_size=0.2, random_state=1)
## Printing the shape of train and test sets
print(X_train_n.shape)
print(y_train_n.shape)
print(X_test_n.shape)
print(y_test_n.shape)

(6744, 250)
(6744,)
(1687, 250)
(1687,)

y_train_n = np.array(y_train_n.values.tolist())
y_test_n = np.array(y_test_n.values.tolist())

from tensorflow.keras.utils import to_categorical

print("Value of train labels before encoding:", y_train_n[0])
print("Value of test labels before encoding:", y_test_n[0])

y_train_nn = to_categorical(y_train_n)
y_test_nn = to_categorical(y_test_n)

print("Shape of y_train_nn:", y_train_nn.shape)
print("One hot encoded value of y_train_nn:", y_train_nn[0])
print("Shape of y_test_nn:", y_test_nn.shape)
print("One hot encoded value of y_test_nn:", y_test_nn[0])

Value of train labels before encoding: 2
Value of test labels before encoding: 39
Shape of y_train_nn: (6744, 68)
One hot encoded value of y_train_nn: [0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]
Shape of y_test_nn: (1687, 68)
One hot encoded value of y_test_nn: [0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.
 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 1. 0. 0. 0. 0. 0. 0. 0. 0.
 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]

nn_model = Sequential()
nn_model.add(Dense(500, input_shape=(len(X_train_n[0]),), activation='relu'))
nn_model.add(Dropout(0.5))
nn_model.add(Dense(200, activation='relu'))
nn_model.add(Dropout(0.3))
nn_model.add(Dense(100, activation='relu'))
nn_model.add(Dropout(0.2))
nn_model.add(Dense(68, activation='softmax', name='Output'))

sgd = SGD(learning_rate=0.01, decay=1e-6, momentum=0.9, nesterov=True)
nn_model.compile(loss='categorical_crossentropy', optimizer=sgd, metrics=['accuracy'])

nn_model.summary()

Model: "sequential_3"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
dense_6 (Dense)              (None, 500)               125500    
_________________________________________________________________
dropout_6 (Dropout)          (None, 500)               0         
_________________________________________________________________
dense_7 (Dense)              (None, 200)               100200    
_________________________________________________________________
dropout_7 (Dropout)          (None, 200)               0         
_________________________________________________________________
dense_8 (Dense)              (None, 100)               20100     
_________________________________________________________________
dropout_8 (Dropout)          (None, 100)               0         
_________________________________________________________________
Output (Dense)               (None, 68)                6868      
=================================================================
Total params: 252,668
Trainable params: 252,668
Non-trainable params: 0
_________________________________________________________________

hist = nn_model.fit(X_train_n, y_train_nn, validation_split = 0.2, epochs=50, batch_size=32, verbose=1)

Epoch 1/50
169/169 [==============================] - 1s 4ms/step - loss: 2.8105 - accuracy: 0.4560 - val_loss: 2.3881 - val_accuracy: 0.4722
Epoch 2/50
169/169 [==============================] - 1s 4ms/step - loss: 2.3077 - accuracy: 0.5186 - val_loss: 2.1417 - val_accuracy: 0.5456
Epoch 3/50
169/169 [==============================] - 1s 4ms/step - loss: 2.1352 - accuracy: 0.5385 - val_loss: 2.0453 - val_accuracy: 0.5441
Epoch 4/50
169/169 [==============================] - 1s 4ms/step - loss: 2.0721 - accuracy: 0.5379 - val_loss: 2.0064 - val_accuracy: 0.5456
Epoch 5/50
169/169 [==============================] - 1s 4ms/step - loss: 2.0231 - accuracy: 0.5414 - val_loss: 1.9753 - val_accuracy: 0.5545
Epoch 6/50
169/169 [==============================] - 1s 3ms/step - loss: 1.9808 - accuracy: 0.5488 - val_loss: 1.9299 - val_accuracy: 0.5745
Epoch 7/50
169/169 [==============================] - 1s 4ms/step - loss: 1.9146 - accuracy: 0.5594 - val_loss: 1.8808 - val_accuracy: 0.5723
Epoch 8/50
169/169 [==============================] - 1s 4ms/step - loss: 1.8734 - accuracy: 0.5605 - val_loss: 1.8412 - val_accuracy: 0.5804
Epoch 9/50
169/169 [==============================] - 1s 4ms/step - loss: 1.8281 - accuracy: 0.5705 - val_loss: 1.8073 - val_accuracy: 0.5812
Epoch 10/50
169/169 [==============================] - 1s 3ms/step - loss: 1.7942 - accuracy: 0.5742 - val_loss: 1.8062 - val_accuracy: 0.5671
Epoch 11/50
169/169 [==============================] - 1s 3ms/step - loss: 1.7621 - accuracy: 0.5702 - val_loss: 1.7553 - val_accuracy: 0.5908
Epoch 12/50
169/169 [==============================] - 1s 3ms/step - loss: 1.7306 - accuracy: 0.5783 - val_loss: 1.7275 - val_accuracy: 0.5893
Epoch 13/50
169/169 [==============================] - 1s 3ms/step - loss: 1.6939 - accuracy: 0.5813 - val_loss: 1.7186 - val_accuracy: 0.5834
Epoch 14/50
169/169 [==============================] - 1s 4ms/step - loss: 1.6560 - accuracy: 0.5872 - val_loss: 1.6955 - val_accuracy: 0.5886
Epoch 15/50
169/169 [==============================] - 1s 3ms/step - loss: 1.6253 - accuracy: 0.5881 - val_loss: 1.6779 - val_accuracy: 0.5960
Epoch 16/50
169/169 [==============================] - 1s 3ms/step - loss: 1.5995 - accuracy: 0.5957 - val_loss: 1.6518 - val_accuracy: 0.5967
Epoch 17/50
169/169 [==============================] - 1s 4ms/step - loss: 1.5755 - accuracy: 0.5931 - val_loss: 1.6431 - val_accuracy: 0.5827
Epoch 18/50
169/169 [==============================] - 1s 4ms/step - loss: 1.5489 - accuracy: 0.5980 - val_loss: 1.6169 - val_accuracy: 0.5930
Epoch 19/50
169/169 [==============================] - 1s 4ms/step - loss: 1.5275 - accuracy: 0.6020 - val_loss: 1.6177 - val_accuracy: 0.5841
Epoch 20/50
169/169 [==============================] - 1s 4ms/step - loss: 1.5201 - accuracy: 0.6078 - val_loss: 1.6139 - val_accuracy: 0.5982
Epoch 21/50
169/169 [==============================] - 1s 3ms/step - loss: 1.4840 - accuracy: 0.6082 - val_loss: 1.6144 - val_accuracy: 0.5990
Epoch 22/50
169/169 [==============================] - 1s 3ms/step - loss: 1.4695 - accuracy: 0.6122 - val_loss: 1.5907 - val_accuracy: 0.5938
Epoch 23/50
169/169 [==============================] - 1s 3ms/step - loss: 1.4506 - accuracy: 0.6102 - val_loss: 1.5865 - val_accuracy: 0.5975
Epoch 24/50
169/169 [==============================] - 1s 3ms/step - loss: 1.4331 - accuracy: 0.6130 - val_loss: 1.5785 - val_accuracy: 0.6034
Epoch 25/50
169/169 [==============================] - 1s 3ms/step - loss: 1.4173 - accuracy: 0.6154 - val_loss: 1.5850 - val_accuracy: 0.5938
Epoch 26/50
169/169 [==============================] - 1s 4ms/step - loss: 1.4125 - accuracy: 0.6185 - val_loss: 1.5750 - val_accuracy: 0.5901
Epoch 27/50
169/169 [==============================] - 1s 3ms/step - loss: 1.3850 - accuracy: 0.6250 - val_loss: 1.5908 - val_accuracy: 0.5960
Epoch 28/50
169/169 [==============================] - 1s 4ms/step - loss: 1.3714 - accuracy: 0.6285 - val_loss: 1.5788 - val_accuracy: 0.5967
Epoch 29/50
169/169 [==============================] - 1s 4ms/step - loss: 1.3587 - accuracy: 0.6252 - val_loss: 1.5739 - val_accuracy: 0.6027
Epoch 30/50
169/169 [==============================] - 1s 3ms/step - loss: 1.3385 - accuracy: 0.6330 - val_loss: 1.5706 - val_accuracy: 0.6019
Epoch 31/50
169/169 [==============================] - 1s 3ms/step - loss: 1.3291 - accuracy: 0.6360 - val_loss: 1.5816 - val_accuracy: 0.5938
Epoch 32/50
169/169 [==============================] - 1s 3ms/step - loss: 1.3183 - accuracy: 0.6365 - val_loss: 1.5768 - val_accuracy: 0.5990
Epoch 33/50
169/169 [==============================] - 1s 4ms/step - loss: 1.2965 - accuracy: 0.6417 - val_loss: 1.6164 - val_accuracy: 0.6086
Epoch 34/50
169/169 [==============================] - 1s 4ms/step - loss: 1.2721 - accuracy: 0.6458 - val_loss: 1.6052 - val_accuracy: 0.6034
Epoch 35/50
169/169 [==============================] - 1s 3ms/step - loss: 1.2545 - accuracy: 0.6539 - val_loss: 1.5798 - val_accuracy: 0.6034
Epoch 36/50
169/169 [==============================] - 1s 3ms/step - loss: 1.2563 - accuracy: 0.6497 - val_loss: 1.6018 - val_accuracy: 0.5982
Epoch 37/50
169/169 [==============================] - 1s 4ms/step - loss: 1.2555 - accuracy: 0.6545 - val_loss: 1.5913 - val_accuracy: 0.6123
Epoch 38/50
169/169 [==============================] - 1s 3ms/step - loss: 1.2266 - accuracy: 0.6558 - val_loss: 1.5898 - val_accuracy: 0.6027
Epoch 39/50
169/169 [==============================] - 1s 3ms/step - loss: 1.2055 - accuracy: 0.6578 - val_loss: 1.6256 - val_accuracy: 0.6145
Epoch 40/50
169/169 [==============================] - 1s 4ms/step - loss: 1.1962 - accuracy: 0.6612 - val_loss: 1.5880 - val_accuracy: 0.6056
Epoch 41/50
169/169 [==============================] - 1s 4ms/step - loss: 1.1853 - accuracy: 0.6665 - val_loss: 1.5971 - val_accuracy: 0.6123
Epoch 42/50
169/169 [==============================] - 1s 4ms/step - loss: 1.1802 - accuracy: 0.6688 - val_loss: 1.6038 - val_accuracy: 0.6056
Epoch 43/50
169/169 [==============================] - 1s 3ms/step - loss: 1.1539 - accuracy: 0.6747 - val_loss: 1.6241 - val_accuracy: 0.6064
Epoch 44/50
169/169 [==============================] - 1s 3ms/step - loss: 1.1617 - accuracy: 0.6727 - val_loss: 1.6091 - val_accuracy: 0.5997
Epoch 45/50
169/169 [==============================] - 1s 4ms/step - loss: 1.1425 - accuracy: 0.6669 - val_loss: 1.6555 - val_accuracy: 0.6101
Epoch 46/50
169/169 [==============================] - 1s 4ms/step - loss: 1.1218 - accuracy: 0.6827 - val_loss: 1.6455 - val_accuracy: 0.6019
Epoch 47/50
169/169 [==============================] - 1s 4ms/step - loss: 1.1108 - accuracy: 0.6801 - val_loss: 1.6339 - val_accuracy: 0.6056
Epoch 48/50
169/169 [==============================] - 1s 3ms/step - loss: 1.1075 - accuracy: 0.6830 - val_loss: 1.6419 - val_accuracy: 0.6027
Epoch 49/50
169/169 [==============================] - 1s 4ms/step - loss: 1.1000 - accuracy: 0.6803 - val_loss: 1.6191 - val_accuracy: 0.5967
Epoch 50/50
169/169 [==============================] - 1s 4ms/step - loss: 1.0865 - accuracy: 0.6840 - val_loss: 1.6411 - val_accuracy: 0.6086

print('Test accuracy is:')
nn_model.evaluate(X_test_n,y_test_nn)

Test accuracy is:
53/53 [==============================] - 0s 2ms/step - loss: 1.6322 - accuracy: 0.6242

[1.6322323083877563, 0.6241849660873413]

nn_test_acc = hist.history['accuracy']
nn_val_acc = hist.history['val_accuracy']
nn_test_loss = hist.history['loss']
nn_val_loss = hist.history['val_loss']

epochs = range(1, len(nn_test_acc) + 1)

plt.plot(epochs, nn_test_acc, 'ro', label='Training accuracy')
plt.plot(epochs, nn_val_acc, 'g', label='Validation accuracy')
plt.title('Training and validation accuracy')
plt.legend()

plt.figure()

plt.plot(epochs, nn_test_loss, 'ro', label='Training loss')
plt.plot(epochs, nn_val_loss, 'g', label='Validation loss')
plt.title('Training and validation loss')
plt.legend()

plt.show()